The DOI matches its Crossref record; 137 of the 148 references it cites were found in a registry or an index; none of the 80 sources its claims cite is checked for support yet.
Missing data in multi-omics integration: Recent advances through artificial intelligence
Open accessAbstract
Biological systems function through complex interactions between various ‘omics (biomolecules), and a more complete understanding of these systems is only possible through an integrated, multi-omic perspective. This has presented the need for the development of integration approaches that are able to capture the complex, often non-linear, interactions that define these biological systems and are adapted to the challenges of combining the heterogenous data across ‘omic views. A principal challenge to multi-omic integration is missing data because all biomolecules are not measured in all samples. Due to either cost, instrument sensitivity, or other experimental factors, data for a biological sample may be missing for one or more ‘omic techologies. Recent methodological developments in artificial intelligence and statistical learning have greatly facilitated the analyses of multi-omics data, however many of these techniques assume access to completely observed data. A subset of these methods incorporate mechanisms for handling partially observed samples, and these methods are the focus of this review. We describe recently developed approaches, noting their primary use cases and highlighting each method's approach to handling missing data. We additionally provide an overview of the more traditional missing data workflows and their limitations; and we discuss potential avenues for further developments as well as how the missing data issue and its current solutions may generalize beyond the multi-omics context.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
This review surveys recent AI/ML multi-omics integration methods that explicitly handle partially observed data, either by joint imputation within the model or by optimization masking of missing components. It also summarizes traditional missing-data workflows and discusses limitations, especially sample-size demands, limited coverage of mass-spectrometry omics, and interpretability.
Key findings
- Missing data are pervasive in omics: proteomics can have 20–50% of possible peptide values unquantified, and the human proteome project estimates ~20% of genes yield protein products not detected by mass spectrometry.
- The review groups missing-data-robust integration methods into joint-imputation approaches (e.g., FBM, iMODA, MOFA/MOFA+, BIDIFAC+) and optimization-masking approaches (e.g., TiMEG, COMBI, MVAE, DeepMF, MvNE, DeepIMV, JACA, SUMO, NEMO, MSNE, MONET), across early, middle, and late integration categories.
- Reported performance varies with missingness: FBM prediction RMSE increased with missing percentage, most prominently when both gene expression and methylation had 30% missingness; iMODA power decreased with 50% and 30% missingness; MOFA imputation had lower MSE than feature-wise mean, SoftImpute, and KNN across ~10–80% missingness; DeepMF identified simulated MCAR sample/feature subgroups with perfect accuracy at 10%, 50%, and 70% missingness; SUMO slightly outperformed NEMO except at extreme missingness (80% and greater).
- BIDIFAC+ enables bi-dimensional integration across omics and sample cohorts, but took nearly 24 h to converge on a combined TCGA dataset, and assumes shared structures are present across integrated matrices.
Methods
- Narrative review of AI/ML and statistical-learning approaches for multi-omics integration that accommodate missing data, plus background on traditional missing-data mechanisms (MCAR, MAR, MNAR) and imputation/complete-case workflows.
- Methods are categorized by integration strategy (early/concatenation, middle/transformation-based, late/model-based) and by missing-data handling strategy (joint-imputation vs. optimization-masking).
- Table 1 summarizes each reviewed method’s integration approach, missing-data approach, compatible omics platforms, primary use case, application dataset characteristics, and software availability.
Limitations
- The paper states that many reviewed methods implicitly require relatively large sample sizes, as their application datasets contain hundreds to thousands of observations, which may be prohibitive for rare diseases or small studies.
- The paper states that many methods were motivated by or evaluated on sequencing-based omics, with few considering metabolomics, proteomics, or lipidomics, where mass-spectrometry-specific batch effects, missingness mechanisms, and identification uncertainty may violate model assumptions.
- The paper notes that only FBM and iMODA describe how correlation structure is modeled or affects results among joint-imputation approaches, and that many middle-integration methods are black-box with limited interpretability; evident from the text, this is a review rather than an independent benchmark, so performance claims are drawn from original publications and simulations.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
54 Not checked yet · 36 Own interpretation · 14 No basis given
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“While all these methods have been presented within the context of multi-omics integration, few explicitly require the use multi-omics data.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“many of the middle integration methods discussed here are black-box in nature, as the specific relationships learned by each are often incommunicable.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“However, data integration is one of the primary barriers to effective multi-omics research and reaching the desired holistic view of the biological system being studied (Misra et al., 2019).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Introduction
What the sources say
- Not checked yetcites Misra et al., 2019 source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Recent methodological developments in artificial intelligence and statistical learning have greatly facilitated the analyses of multi-omics data, however many of these techniques assume access to completely observed data.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Abstract
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“A subset of these methods incorporate mechanisms for handling partially observed samples, and these methods are the focus of this review.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Abstract
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“Machine learning (ML) and artificial intelligence (AI) have shown promise in overcoming some of these hindrances (e.g., Mirza et al., 2019; Zhu, 2020) and have gained popularity in recent years.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Introduction
What the sources say
plainconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“However, the observed decreases in predictive performance were observed to be mitigated as the sample size was increased.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Early integration
What the sources say
- Not checked yetcites Fang et al., 2018 source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- The authors state this as an assumption: The FBM model assumes gene-gene and methylation-methylation independence, and imputed values reflect this assumption.
None of the sources cited for this claim have been checked yet.
In the paper
“For example, the human proteome project reports that an estimated ~20% of genes yield protein products that are not detected by mass spectrometry, an analytic platform for protein quantification (Paik et al., 2012; Baker et al., 2017).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Introduction
What the sources say
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“A similar trend is observed for block-level missingness, which is the type of missingness where several omics (columns) are missing across a set of samples (rows).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Middle integration
What the sources say
- Not checked yetcites Lock et al., 2022 source
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“It is not uncommon to have 20–50% of the possible peptide values that are not quantified (Webb-Robertson et al., 2015; Brenes et al., 2019).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
What the sources say
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- It may state two separate claims as one. 41
- Probably not a substantial claim. 30
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Identifying these groupings within latent factors is often a post-hoc process, however, and such domain-relevant groupings are not always identifiable within generated latent factors.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“A principal challenge to multi-omic integration is missing data because all biomolecules are not measured in all samples.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Abstract
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Hinderances to effective integration methods include, but are not limited to, the heterogeneous nature of data and distributional properties across multi-omics datasets, the typically large difference in the number of measured biomolecules compared to number of samples or replicates, and the complex and often noisy nature of biological data.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Introduction
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Due to either cost, instrument sensitivity, or other experimental factors, data for a biological sample may be missing for one or more ‘omic techologies.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Abstract
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
In the paper
“Furthermore, additional complexities including protein isolation and solubilization, sequence ambiguity, and a lack of standards in statistical thresholds and algorithms result in inconsistent detection of large proportions of protein products that are detected, thus limiting the reproducibility of data collection and analysis (Goh and Wong, 2017).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Introduction
What the sources say
- Not checked yetcites Goh and Wong, 2017
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In ‘omics studies, data are more commonly MNAR or MAR, considering that instruments are known to have certain detection thresholds (MNAR) and that biochemical dependencies exist between different biomolecules (MAR).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“If an entire view is missing for one or more samples and must be imputed, any level of similarity specific to that view may be lost.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Multi-omics integration
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Missing data are common in ‘omics data and can arise due to a variety of reasons, such as poor tissue quality, insufficient sample volume, measurement system limitations, budget restrictions, or subject dropout.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“When faced with missing data, these methods thus require a complete case analysis or the implementation of separate imputation steps beforehand.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Multi-omics integration
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“However the authors note that models incorporating gene-gene and methylation-methylation correlations were additionally considered, but these models were associated with a higher computational cost without indicating substantive gains in performance relative to models based on the independence assumption.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Early integration
What the sources say
- Not checked yetcites Fang et al., 2018 source
hedgedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- The authors state this as an assumption: The FBM model assumes gene-gene and methylation-methylation independence, and imputed values reflect this assumption.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Given the prevalence of missing data across different ‘omics, the approach that ML and AI models use in handling missing observations is of great importance when considering multi-omics integration methods.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Introduction
plain- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence.
In the paper
“KNN imputation has been shown to work well for numerical imputation on single datasets (Jadhav et al., 2019), but the algorithm is sensitive to the choice of the number of nearest neighbors (k) and is limited in that it does not utilize the potential structure across the entire set of observations.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
What the sources say
- Not checked yetcites Jadhav et al., 2019 source
boosteddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“While more robust than a complete case analysis since less data are dropped, an available case analysis is still only valid under the MCAR assumption and generates conclusions based on different parts of the data, potentially leading to biased or conflicting inference.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“However, like mean imputation, imputed values from these approaches may distort feature distributions and lead to biased parameter and variability estimates (Lubin et al., 2004; Succop et al., 2004; Webb-Robertson et al., 2015; Lazar et al., 2016; Bramer et al., 2020).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
What the sources say
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“Middle integration applies some transformation to represent a complex combination of the datasets before applying downstream analyses.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Multi-omics integration
What the sources say
- Not checked yetcites Ritchie et al., 2015 source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Missing samples in multi-omics data are rarely MCAR, and thus inference based on a complete case analysis is likely to be biased.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Multi-omics integration
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“Both these methods require complete input, and initial imputation is performed in both cases to accommodate.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Multi-omics integration
What the sources say
- Not checked yetcites Xie et al., 2019 source
- Not checked yetcites Xu et al., 2019
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“These methods rely on an imputation pre-processing step, which could possibly negatively affect the downstream analyses, as the relationships between nodes/samples/biomarkers may be biased toward similarities in imputation.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Multi-omics integration
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Integration techniques generally assume measurements from the same samples on several markers, or features, from several ‘omics-data types, which are more generally referred to as data “views”.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Multi-omics integration
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“While this simple approach preserves feature means and grants access to the use of more data, it does not preserve relationships among variables and may lead to an underestimate of variability (Gelman and Hill, 2006).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
What the sources say
- Not checked yetcites Gelman and Hill, 2006 source
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“In metabolomics, limited coverage of the known metabolome increases the risk of overlooking the metabolomic response of interest in targeted metabolomic analyses.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Introduction
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
In the paper
“Early integration is primarily characterized by an initial concatenation of the features across all measured ‘omics, followed by application of methods appropriate for analyzing the resulting high-dimensional dataset.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Multi-omics integration
What the sources say
- Not checked yetcites Ritchie et al., 2015 source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“Late integration analyzes each dataset separately and introduces a model or algorithm that combines the outputs of each individual analysis.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Multi-omics integration
What the sources say
- Not checked yetcites Ritchie et al., 2015 source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“Peptides may be missing from one or more samples due to a variety of underlying mechanisms (Daly et al., 2008; Karpievitch et al., 2009), which can result in non-trivial amounts of potential observations being missing.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
What the sources say
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Given the ubiquity of partially observed cases in multi-omics datasets—and that most integrative multi-omics methods are not inherently compatible with partially observed data—imputation methods are often incorporated as a data pre-processing step within the multi-omics analytic pipeline, or alternatively, partially observed samples are dropped and a complete case analysis is performed.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Given the tendency of neural networks to overfit to the training data, it is possible that any bias in the imputation of the training data—which is likely due to the common violation of MAR in ‘omics data—will be erroneously exploited by the network.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Multi-omics integration
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“The missing percentages considered ranged between ~10–~80%, and all methods demonstrated increases to mean squared error of prediction with increased missing percentage.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Middle integration
What the sources say
- Not checked yetcites Argelaguet et al., 2018 source
boosteddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“This approach, while convenient, results in a decrease in sample size, throws away useful data, is only valid under the MCAR assumption and can otherwise lead to biased estimates (Fang et al., 2018; Hawinkel et al., 2020).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
What the sources say
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Each of the previously described imputation methods have generally been developed within the context of a single dataset, however multi-omics data are uniquely characterized by several separate, yet related, datasets. This aspect of multi-omics has therefore motivated the development of several newer approaches that leverage information across different ‘omics datasets to inform imputations.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Background > Missing data and imputation
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“The power of the iMODA framework to detect the effects of interest has been reported to vary according to the proportion of data that are missing. The authors considered cases where 50% and 30% of data were missing, finding that power decreases with increased missingness.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Early integration
What the sources say
- Not checked yetcites Lin et al., 2020 source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The known causal hierarchy across biomolecules is that genotype affects omic expression, which influences phenotype.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Related to this point, many of the methods described here were motivated by or evaluated exclusively on sequencing-based omics platforms (e.g., transcriptomics, genomics, and epigenomics) and few considered the fuller breadth of omics platforms that includes metabolomics, proteomics, lipidomics, etc.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Contrasting with the sequencing-based ‘omics used by many of the methods reviewed in this work, data of other ‘omics are typically generated by a mass spectrometer, an instrument whose use introduces several nuances to the data which may mean integration model assumptions reasonably made for sequencing-based data may not hold for mass spectrometry-based data.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The datasets considered by these approaches contain hundreds, if not thousands, of observations, sample sizes that may be practically prohibitive due to cost, biospecimen availability, or both.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“Notably, there is no constraint on the number of detected modules, nor is there an upper bound on module size.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Late integration
What the sources say
- Not checked yetcites Rappoport et al., 2020 source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
In the paper
“However, Lock et al. (2022) note that BIDIFAC+ assumes that shared structures are present across the integrated matrices, an assumption that may be violated in practice.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
What the sources say
- Not checked yetcites Lock et al., 2022 source
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
- The authors state this as an assumption: BIDIFAC+ assumes that shared structures are present across the integrated matrices.
In the paper
“This PoE model architecture is robust to partially observed samples, combining only the ‘omic-specific encodings of each partial sample's observed ‘omics during model training.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Middle integration
What the sources say
- Not checked yetcites Wu and Goodman (2018)
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- The authors state this as an assumption: MVAE assumes that data views are independent from one another, given a shared latent representation.
None of the sources cited for this claim have been checked yet.
In the paper
“As well, the range of adjusted rand index values (the performance metric used) was typically observed to be narrower for SUMO, suggesting less variability in expected performance.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Middle integration
What the sources say
hedgedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“The authors circumvent this issue by proposing a heuristic algorithm that identifies a local maximum and recommend using the best solution from repeated runs of the algorithm to better approximate the globally maximal solution.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Late integration
What the sources say
- Not checked yetcites Rappoport et al., 2020 source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may state two separate claims as one.
In the paper
“However, it should be noted that these evaluations were completed in a simulated data setting in which data were simulated as missing completely at random (MCAR) and ground truth feature subgroups were generated with simple discrimination boundaries.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Middle integration
What the sources say
- Not checked yetcites Chen et al., 2019 source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“These methods also commonly investigate differences in “early”, “middle”, and “late” fusion and work has even been done to address the view-missing and view-heterogeneity problems (Nagrani et al., 2018; Wang et al., 2020).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
What the sources say
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The criteria of sets of features from distinct views across a shared set of samples is applicable to any domain where the combination of different data sources provides a more complete representation of the system under investigation.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“On the other hand, though dependencies exist between different ‘omics data, important differences between ‘omics must also be accounted for in developed integrative approaches.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“In untargeted metabolomics and lipidomics false discovery rates are difficult to compute and largely not accurate (Stein, 1994; Jeong et al., 2011; Matsuda et al., 2013; Kim and Zhang, 2014).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
What the sources say
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“This algorithm incorporates partially observed samples by representing their missing components as edgeless nodes within their corresponding edge-weighted graphs.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Late integration
What the sources say
- Not checked yetcites Rappoport et al., 2020 source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“These cases present the need for methods that generate reliable inference despite smaller sample sizes and avoid overfitting to the data.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
plaindirectn = 16- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“many of the deep-learning architectures mentioned in this review have similarities to ones used in analysis of digital media that try to combine multiple modalities to improve predictive performance of video content (Carreira and Zisserman, 2017; Afouras et al., 2019).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
What the sources say
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“For example, a study to identify biomarkers for Snyder Robinson syndrome, a rare neurological disorder, obtained data on 16 patients, only three of which presented with the disorder of interest (Abela et al., 2015).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
What the sources say
- Not checked yetcites Abela et al., 2015 source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“manual verification of metabolites following identification scoring is required for shotgun metabolomics, as scoring methods are much less mature than for other ‘omics data types (Stein and Scott, 1994; Samaraweera et al., 2018).”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
What the sources say
- Not checked yetcites Stein and Scott, 1994
- Not checked yetcites Samaraweera et al., 2018 source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We describe this requirement as “implicit” given that it is rarely, if ever, stated as a limitation, but instead implied based on the sample sizes of the datasets that these methods were successfully applied to.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“Using the area under the receiver-operator curve (AUROC) as a performance metric, the authors found that the performance of DeepIMV diminished with increased missingness but performed as well as or better than the compared approaches.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Integrative approaches for partially observed multi-omics > Middle integration
What the sources say
- Not checked yetcites Lee and van der Schaar, 2021
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“As well, they noted that BIDIFAC+ took nearly 24 h to reach convergence when fitted to a combined TCGA dataset, and thus the method may not scale well with larger aggregations of datasets.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
What the sources say
- Not checked yetcites Lock et al., 2022 source
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
In the paper
“Kok et al. (2018) notes, for example, that one may need around 19 samples per group to detect a fold change of 1.5 in microRNA data at 80% power and false discovery rate <10%.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
What the sources say
- Not checked yetcites Kok et al., 2018 source
hedgedn = 19- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Insights from the ‘omics domain, where some of these issues are most salient, could transfer to other domains.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The methodological choices inherent to the ‘omics domain are usually the ones that offer the ability to constrain or elucidate relationships between biomarkers.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“the groupings implicit within the latent factors generated by these models may be contextualized according to domain-specific knowledge (e.g., shared molecular pathways), thus increasing their interpretability and allowing one to infer underlying relationships.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Some common choices are the use of principal-component-like representations that might identify important groups of biomarkers, and the injection of prior knowledge through techniques that constrain the relationships between features such as directed edges in graph representations, hard parameter constraints, or carefully chosen priors in”
- Also stated in: § Discussion · plain
“the relationships between features such as directed edges in graph representations, hard parameter constraints, or carefully chosen priors in Bayesian frameworks.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
- Also stated in: § Discussion · plain
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“incorporating domain knowledge may be key toward bridging the interpretability gap apparent across these methods.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“If the generation of these latent factors were instead guided by domain knowledge, resulting groupings may be more likely to hold relevance and rely less on post-hoc analyses for contextualization”
- Also stated in: § Discussion · hedged
“instead guided by domain knowledge, resulting groupings may be more likely to hold relevance and rely less on post-hoc analyses for contextualization within the scientific domain.”
§ Missing data in multi-omics integration: Recent advances through artificial intelligence > Discussion
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.
- Also stated in: § Discussion · hedged
The paper relies on this without support.
Existing multi-omics data integration review articles either do not address or only briefly mention methods for handling missing observations.
“Despite the importance of considering methods for handling missing observations, this topic is either not addressed or is only briefly mentioned in existing multi-omics data integration review articles.”
The authors state this as an assumption.
The iBAG approach assumes completely observed data.
“The iBAG approach assumes completely observed data, and to remedy this, the FBM approach introduces a third, imputation layer that models missing gene expression and methylation data through multivariate normal distributions.”
The authors state this as an assumption.
The FBM model assumes gene-gene and methylation-methylation independence, and imputed values reflect this assumption.
“Also worth noting is that the FBM model assumes gene-gene and methylation-methylation independence and imputed values reflect this assumption accordingly.”
The authors state this as an assumption.
MVAE assumes that data views are independent from one another, given a shared latent representation.
“One key assumption made by this approach is that data views are independent from one another, given a shared latent representation.”
The authors state this as an assumption.
BIDIFAC+ assumes that shared structures are present across the integrated matrices.
“Lock et al. (2022) note that BIDIFAC+ assumes that shared structures are present across the integrated matrices, an assumption that may be violated in practice.”
The paper relies on this without support.
The known causal hierarchy across biomolecules is that genotype affects omic expression, which influences phenotype.
“The proposed framework is aligned with the known causal hierarchy across these biomolecules, i.e., genotype affects ‘omic expression which influences phenotype.”
The paper relies on this without support.
Gene expression is known to be affected by changes to the DNA sequence.
“For example, gene expression is known to be affected by changes to the DNA sequence.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“Although these methods are simple and easy to implement, they can introduce bias and reduce data variability (Flores et al., 2023).”
From Navigating the landscape of multimodal AI in medicine: A scoping review on technical challenges and clinical applications (2025) · cites this paper as Flores et al., 2023
Citations
Cites
Maximum Likelihood from Incomplete Data Via the EM Algorithm
Dempster, A. P., Laird, N. M., & Rubin, D. B. (1977) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.2517-6161.1977.tb01600.x · Cited by 50,444 · cited 2×
“Missing ‘omics data are accommodated through an expectation-maximization (EM) approach (Dempster et al., 1977) for parameter estimation.”
MethodOn Information and Sufficiency
Kullback, S., & Leibler, R. A. (1951) · The Annals of Mathematical Statistics · doi:10.1214/aoms/1177729694 · Cited by 20,193
Auto-Encoding Variational Bayes
Kingma, D. P., & Welling, M. (2013) · arXiv · doi:10.48550/arxiv.1312.6114 · Cited by 20,067
mice: Multivariate Imputation by Chained Equations in R
Buuren, S. V., & Groothuis-Oudshoorn, K. (2011) · Journal of Statistical Software · doi:10.18637/jss.v045.i03 · Cited by 13,870
Data Analysis Using Regression and Multilevel/Hierarchical Models
Gelman, A., & Hill, J. (2006) · doi:10.1017/cbo9780511790942 · Cited by 13,597
RUBIN, D. B. (1976) · Biometrika · doi:10.1093/biomet/63.3.581 · Cited by 9,925
Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset
Carreira, J., & Zisserman, A. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.502 · Cited by 9,570 · cited 1×
“For example, many of the deep-learning architectures mentioned in this review have similarities to ones used in analysis of digital media that try to combine multiple modalities to improve predictive performance of video content (Carreira and Zisserman, 2017; Afouras et al., 2019).”
Hanrahan, G. (2008) · CRC Press eBooks · doi:10.1201/9781420067972-10 · Cited by 8,196
(2010) · Cited by 7,034
Missing Value Estimation Methods for DNA Microarrays
Troyanskaya, O., Cantor, M., Sherlock, G., Brown, P., Hastie, T., Tibshirani, R., Botstein, D., & Altman, R. B. (2001) · Bioinformatics · doi:10.1093/bioinformatics/17.6.520 · Cited by 4,230
Research works printed but not linked: 3
- Achille, 2017, Emergence of invariance and disentanglement in deep representations, J. Machine Learn. Res., 19, 1, 10.48550/arxiv.1706.01350 · DOI
- A variational information bottleneck approach to multi-omics data integration,15131521 LeeC. van der SchaarM. International Conference on Artificial Intelligence and Statistics2021
- Sindhwani, 2005, A co-regularization approach to semi-supervised learning with multiple views,
Cited by
Sorted by citations
Deep learning-based approaches for multi-omics data integration and analysis
Ballard, J. L., Wang, Z., Li, W., Shen, L., & Long, Q. (2024) · BioData Mining · doi:10.1186/s13040-024-00391-z · Cited by 301
“The former results in a reduced sample size while not allowing the full usage of all information contained in the dataset, and the samples with missing values may not be a random subset of the data [101].”
BackgroundBaião, A. R., Cai, Z., Poulos, R. C., Robinson, P. J., Reddel, R. R., Zhong, Q., Vinga, S., & Gonçalves, E. (2025) · Briefings in Bioinformatics · doi:10.1093/bib/bbaf355 · Cited by 274
Cembrowska-Lech, D., Krzemińska, A., Miller, T. I., Nowakowska, A., Adamski, C., Radaczyńska, M., Mikiciuk, G., & Mikiciuk, M. (2023) · Biology · doi:10.3390/biology12101298 · Cited by 193
“Moreover, AI and machine learning techniques can be leveraged to analyze the large and complex datasets generated by these methods, revealing hidden patterns and predictive models of plant phenotypes [56–58].”
BackgroundRevolutionizing multi‐omics analysis with artificial intelligence and data processing
Yetgin, A. (2025) · Quantitative Biology · doi:10.1002/qub2.70002 · Cited by 174
“Due to the size and complexity of the data produced by various omics technologies, AI and data processing are essential elements in multi ‐ omics analysis [16].”
Acharya, D., & Mukhopadhyay, A. (2024) · Briefings in Functional Genomics · doi:10.1093/bfgp/elae013 · Cited by 148
Schouten, D., Nicoletti, G., Dille, B., Chia, C., Vendittelli, P., Schuurmans, M., Litjens, G., & Khalili, N. (2025) · Medical Image Analysis · doi:10.1016/j.media.2025.103621 · Cited by 162
Frontiers in pancreatic cancer on biomarkers, microenvironment, and immunotherapy
Yu, B., Shao, S., & Ma, W. (2024) · Cancer Letters · doi:10.1016/j.canlet.2024.217350 · Cited by 119
Cardiac Fibrosis in the Multi-Omics Era: Implications for Heart Failure
Ghazal, R., Wang, M., Liu, D., Tschumperlin, D. J., & Pereira, N. L. (2025) · Circulation Research · doi:10.1161/circresaha.124.325402 · Cited by 128
Biomarkers of aging in frailty and age-associated disorders: State of the art and future perspective
Salvioli, S., Basile, M. S., Bencivenga, L., Carrino, S., Conte, M., Damanti, S., LORENZO, R. D., Fiorenzato, E., Gialluisi, A., Ingannato, A., et al. (2023) · Ageing Research Reviews · doi:10.1016/j.arr.2023.102044 · Cited by 91
Multi-omics integration analysis: Tools and applications in environmental toxicology
Shi, C., Cheng, L., Yu, Y., Chen, S., Dai, Y., Yang, J., Zhang, H., Chen, J., & Geng, N. (2024) · Environmental Pollution · doi:10.1016/j.envpol.2024.124675 · Cited by 85
Sen, P., & Orešič, M. (2023) · Metabolites · doi:10.3390/metabo13070855 · Cited by 75
Amin, A., Zaman, W., & Park, S. (2025) · Genes · doi:10.3390/genes16070809 · Cited by 82
Artificial intelligence and multi-omics in pharmacogenomics: a new era of precision medicine
Zack, M., Stupichev, D. N., Moore, A. J., Slobodchikov, I. D., Sokolov, D. G., Trifonov, I. F., & Gobbs, A. (2025) · Mayo Clinic Proceedings: Digital Health · doi:10.1016/j.mcpdig.2025.100246 · Cited by 70
“23,24 It recognizes that genotype alone does not capture dynamic processes such as gene regulation, post-translational modifications, or shifts in metabolism.”
Ganie, S. A., McMulkin, N., & Devoto, A. (2024) · Plant Cell & Environment · doi:10.1111/pce.14855 · Cited by 71
Multi-omics data integration and analysis pipeline for precision medicine: Systematic review
Abdelaziz, E. H., Ismail, R. M., Mabrouk, M. S., & Amin, E. (2024) · Computational Biology and Chemistry · doi:10.1016/j.compbiolchem.2024.108254 · Cited by 67
Machine learning for the advancement of genome-scale metabolic modeling
Kundu, P., Beura, S., Mondal, S., Das, A. K., & Ghosh, A. (2024) · Biotechnology Advances · doi:10.1016/j.biotechadv.2024.108400 · Cited by 53
Dave, J. K., Jagana, V., Janoštiak, R., & Bisserier, M. (2023) · Journal of Translational Medicine · doi:10.1186/s12967-023-04339-5 · Cited by 50
“However, with recent artificial intelligence developments and publicly accessible datasets, multiomics data analysis has greatly improved, assuming access to all observed data following sample testing [131, 135].”
BackgroundIntegrative Multi-Omics and Artificial Intelligence: A New Paradigm for Systems Biology
Kant, S., Deepika, & Roy, S. (2025) · OMICS A Journal of Integrative Biology · doi:10.1177/15578100251392371 · Cited by 61
“Methods used for imputation span low-level techniques such as mean or median imputation, to hot-deck, traditional ML, to advanced deep learning (DL) models suited for high-dimensional omics data (Flores et al., 2023; Voillet et al., 2016).”
A machine learning and deep learning-based integrated multi-omics technique for leukemia prediction
Abbasi, E. Y., Deng, Z., Arain, Q. A., Khan, A., Shaikh, A., Reshan, M. S. A., Sulaiman, A. A., & Alshahrani, H. M. (2024) · Heliyon · doi:10.1016/j.heliyon.2024.e25369 · Cited by 49
“Only features with more than 20 % and less than 50 % of the data missing score were included [61].”
BackgroundBernardini, M., Doinychko, A., Romeo, L., Frontoni, E., & Amini, M.-R. (2023) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2023.107188 · Cited by 42
PathIntegrate: Multivariate modelling approaches for pathway-based multi-omics data integration
Wieder, C., Cooke, J., Frainay, C., Poupin, N., Bowler, R. P., Jourdan, F., Kechris, K. J., Lai, R., & Ebbels, T. M. D. (2024) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1011814 · Cited by 44
“However, further work is required to develop multi-omics integration methods that can handle samples with one or more omics missing [70] (missing not at random [68]).”
Machine learning for extracellular vesicles enables diagnostic and therapeutic nanobiotechnology
Tiwari, A. K., Widodo, Krisnawati, D. I., Tzou, K.-Y., & Kuo, T. (2026) · Journal of Nanobiotechnology · doi:10.1186/s12951-025-03952-4 · Cited by 55
Yoon, J. H., Lee, H., Kwon, D., Lee, D., Lee, S., Cho, E., Kim, J., & Kim, D. (2024) · Brain Communications · doi:10.1093/braincomms/fcae265 · Cited by 43
“79,80 As many trials using genome-wide association studies and multi-omics data to detect Alzheimer’s disease risk genes and drug targets are not successful, attempts including drug repurposing have been made to apply statistics and machine/deep learning to multi-omics to gain new insights.”
BackgroundMultimodal Integration in Health Care: Development With Applications in Disease Management
Hao, Y., Cheng, C., Li, J., Li, H., Di, X., Zeng, X., Jin, S., Han, X., Liu, C., Wang, Q., et al. (2025) · Journal of Medical Internet Research · doi:10.2196/76557 · Cited by 44
“Inconsistent data collection practices, missing entries, and recording errors can compromise model reliability, necessitating robust standardization protocols [84].”
Chen, Z., Liu, A., Zhang, X., Cao, X., Ding, Z., Sha, Q., Shen, H., Deng, H., & Zhou, W. (2024) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2024.108058 · Cited by 39
Integrating multi-omics and machine learning for disease resistance prediction in legumes
Mohamedikbal, S., Al-Mamun, H. A., Bestry, M., Batley, J., & Edwards, D. (2025) · Theoretical and Applied Genetics · doi:10.1007/s00122-025-04948-2 · Cited by 33
Ouhmouk, M., Baichoo, S., & Abık, M. (2025) · Informatics in Medicine Unlocked · doi:10.1016/j.imu.2025.101679 · Cited by 32
Elrashedy, A., Mousa, W. S., Nayel, M., Salama, A. A., Zaghawa, A. A., Elsify, A. M., & Hasan, M. E. (2025) · Virology Journal · doi:10.1186/s12985-025-02640-x · Cited by 25
Liu, Z., & Park, T. (2024) · Frontiers in Genetics · doi:10.3389/fgene.2024.1488683 · Cited by 25
Computational Methods for Data Integration and Imputation of Missing Values in Omics Datasets
Schumann, Y., Gocke, A., & Neumann, J. E. (2024) · PROTEOMICS · doi:10.1002/pmic.202400100 · Cited by 31
“In particular, this work focuses on the integration of multiple cohorts from the same omic -type and the reader is referred to the reviews in [1–3] for an overview of multi-omic data integration and to [4] for an integrated consideration of multi-omics with missing values.”
Multi-Omics Integration in Nephrology: Advances, Challenges, and Future Directions
Saliba, A., Du, Y., Feng, T., & Garmire, L. X. (2024) · Seminars in Nephrology · doi:10.1016/j.semnephrol.2025.151584 · Cited by 32
“Readers are referred to more details in the review of Flores et al. 28”
Computational Metagenomics: State of the Art
Pita-Galeana, M. A., Ruhle, M., López-Vázquez, L., de Anda‐Jáuregui, G., & Lemus, E. H. (2025) · International Journal of Molecular Sciences · doi:10.3390/ijms26189206 · Cited by 28
“ML and AI methods such as k-nearest neighbors (KNN) imputation, random forest imputation, and expectation-maximization algorithms better capture relationships among features [160].”
InfluentialMoving towards genome-wide data integration for patient stratification with Integrate Any Omics
Ma, S., Zeng, A. G. X., Haibe‐Kains, B., Goldenberg, A., Dick, J. E., & Wang, B. (2025) · Nature Machine Intelligence · doi:10.1038/s42256-024-00942-3 · Cited by 33
“Advanced integrative methods to address the missing data issue can be classified into two categories: joint imputation or optimization masking 22 .”
Ameliorating microalgal OMEGA production using omics platforms
Mariam, I., Bettiga, M., Rova, U., Christakopoulos, P., Μάτσακας, Λ., & Patel, A. (2024) · Trends in Plant Science · doi:10.1016/j.tplants.2024.01.002 · Cited by 25
Human‐based new approach methodologies to accelerate advances in nutrition research
Cassotta, M., Cianciosi, D., Elexpuru-Zabaleta, M., Elío, I., Cano, S. S., Giampieri, F., & Battino, M. A. (2024) · Food Frontiers · doi:10.1002/fft2.369 · Cited by 23
Evaluation of normalization strategies for mass spectrometry-based multi-omics datasets
Tseng, C. Y., Salguero, J. A., Breidenbach, J. D., Solomon, E. A., Sanders, C. K., Harvey, T., Thornhill, M. G., Palmisano, S. J., Sasiene, Z. J., Blackwell, B. R., et al. (2025) · Metabolomics · doi:10.1007/s11306-025-02297-1 · Cited by 25
“These technical factors affect omics feature identification, quantification, and missing value gen - eration, varying by omics type and complicate the selection of normalization methods (Brombacher et al., 2025; Flores et al., 2023; Kong et al., 2022).”
The impact of artificial intelligence on biomarker discovery
Javaid, H., Petrescu, C. C., Schmunk, L. J., Monahan, J., O’Reilly, P. G., Garg, M., McGirr, L., Khasawneh, M., Lail, M. A., Ganta, D., et al. (2025) · Emerging Topics in Life Sciences · doi:10.1042/etls20243003 · Cited by 32
Omics big data for crop improvement: Opportunities and challenges
Vasupalli, N., Bhat, J. A., Jain, P., Sri, T., Islam, M. A., Shivaraj, S. M., Singh, S. K., Deshmukh, R., Sonah, H., & Lin, X. (2024) · The Crop Journal · doi:10.1016/j.cj.2024.10.007 · Cited by 21
MARIA: A multimodal transformer model for incomplete healthcare data
Caruso, C. M., Soda, P., & Guarrasi, V. (2025) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2025.110843 · Cited by 30
“Various techniques are available for fusing latent representations from different modalities, and we recommend readers refer to the review [1] for an in-depth exploration of these methods. imputer, which has demonstrated effectiveness in handling missing values in tabular data [5, 6, 7].”
The Immune System and Cellular Senescence: A Complex Interplay in Aging and Disease
Ayoub, M., Jaoude, C. A., Ayoub, M., Hamadé, A., & Rima, M. (2025) · Immunology · doi:10.1111/imm.70036 · Cited by 25
“More accurate and customised immunosenescence assessment systems are being created as a result of the rapid advancement of multi-omics techniques (genomics, transcrip-tomics, proteomics and metabolomics) discussed in Section 6.1, and the incorporation of artificial intelligence [229, 231–232].”
InfluentialRadío, S., Marsico, M. D., Bersani, C., Malinverni, R., Casacuberta, J., Corpetti, C., Cigliano, R. A., & Sanseverino, W. (2024) · EFSA Supporting Publications · doi:10.2903/sp.efsa.2024.en-9086 · Cited by 16
Hsieh, Y. E., Tandon, K., Verbruggen, H., & Nikoloski, Z. (2025) · The ISME Journal · doi:10.1093/ismejo/wraf109 · Cited by 15
“However, metatranscriptomic data are often sparse, with relatively few genes showing considerable expression,which can be influenced by taxon and/or gene–family abundance [53–55].”
Olawade, D. B., Kade, A., Egbon, E., Usman, S. O., Fapohunda, O., Ijiwade, J. O., & Ogbonna, C. E. (2025) · Molecular Genetics and Genomics · doi:10.1007/s00438-025-02314-x · Cited by 21
“Furthermore, integration must include varying sizes and measurement units across distinct omics layers (Flores et al. 2023 ).”
Leveraging Artificial Intelligence for Synergies in Drug Discovery: From Computers to Clinics
Arora, P., Behera, M., Saraf, S. A., & Shukla, R. (2024) · Current Pharmaceutical Design · doi:10.2174/0113816128308066240529121148 · Cited by 15
Key Proteomics Tools for Fundamental and Applied Microalgal Research
Plouviez, M., & Dubreucq, É. (2024) · Proteomes · doi:10.3390/proteomes12020013 · Cited by 13
“In addition, especially for high-throughput techniques, huge datasets are being generated and, therefore, require the use of bioinformatics tools and pipelines to acquire and analyze data [24,31,88–90].”
BackgroundResultInfluentialFeng, J., Liu, Y., Li, K., & Wu, Y. (2025) · International Journal of Pharmaceutics · doi:10.1016/j.ijpharm.2025.125332 · Cited by 12
Zhan, A. (2025) · Biological Diversity · doi:10.1002/bod2.70015 · Cited by 12
Network-based analyses of multiomics data in biomedicine
Kumar, R., Romano, J. D., & Ritchie, M. D. (2025) · BioData Mining · doi:10.1186/s13040-025-00452-x · Cited by 15
“The most prominent in this review were the curse of dimensionality (having far more features than samples in many cases) [68, 69] and the issue of addressing noise and sparsity in some datasets [15, 70].”
Torres-Narvaez, E. S., González, D., Artunduaga-Alvarado, J. A., & Ortega‐Recalde, O. (2025) · Frontiers in Oncology · doi:10.3389/fonc.2025.1584178 · Cited by 16
Emerging Trends in Systems Biology: Multi-Omics Integration and Beyond
Wang, N., Zhang, G., & Li, M. (2024) · Computational Molecular Biology · doi:10.5376/cmb.2024.14.0024 · Cited by 10
Polizel, G. H. G., Fanalli, S. L., Diniz, W. J. S., Cesar, A. S. M., Cônsolo, N. R. B., Fukumasu, H., Cánovas, Á., Fernandes, A. C., Prati, B. C. T., Furlan, É., et al. (2024) · Scientific Reports · doi:10.1038/s41598-024-78965-4 · Cited by 11
“Following the protocol outlined for proteome and metabolome datasets 23 , we chose a soft threshold of 15, resulting in a mean connectivity of 8.40.”
InfluentialUnlocking biological complexity: the role of machine learning in integrative multi-omics
Kumar, R., Ruhel, R., & van Wijnen, A. J. (2024) · Academia Biology · doi:10.20935/acadbiol7428 · Cited by 12
“The synergy between ML and multi-omics holds immense promise for advancing our understanding of biology and improving clinical outcomes [6].”
Navigating multi-omic integration methods for human microbiome research
Muller, E., Bamberger, T., & Borenstein, E. (2026) · Nature Microbiology · doi:10.1038/s41564-026-02328-0 · Cited by 12
“…integration where each omic is first modelled and then models’ outcomes are combined) 42 , (4) the characteristics of the omic data that can be integrated (for example, supported number of omics, data types, distributions and so on) and finally, (5) the algorithmic scale and distribution 35–37 .”
Vo, T. H., & Le, N. Q. K. (2026) · Briefings in Bioinformatics · doi:10.1093/bib/bbag227 · Cited by 12
“These modality-specific differences are widely documented in multi-omics integration studies as major challenges for data harmonization and fusion, particularly due to heterogeneity across data types, high dimensionality, and incomplete modality coverage [40–42].”
InfluentialComputational Strategies for Assessing Adverse Outcome Pathways: Hepatic Steatosis as a Case Study
Ortega-Vallbona, R., Palomino‐Schätzlein, M., Tolosa, L., Benfenati, E., Ecker, G. F., Gozalbes, R., & Serrano‐Candelas, E. (2024) · International Journal of Molecular Sciences · doi:10.3390/ijms252011154 · Cited by 10
“ML models require larger sample sizes for a correct fitting and can suffer from overfitting challenges if dimensionality problems arise from having too many omics variables associated with insufficient samples [117,188,191].”
BackgroundMethodInfluentialFehr, K. B., Mertens, A., Shu, C.-H., Dailey-Chwalibóg, T., Shenhav, L., Allen, L. H., Beggs, M. R., Bode, L., Chooniedass, R., DeBoer, M. D., et al. (2025) · Frontiers in Nutrition · doi:10.3389/fnut.2025.1548739 · Cited by 10
Wettinger, S. B., Karađuzović-Hadžiabdić, K., Attard, R., Farrugia, R. C., Wolford, B. N., Chierici, M., Jurman, G., Alexiou, P., Peñalvo, J. L., Costa, R. S., et al. (2025) · Briefings in Bioinformatics · doi:10.1093/bib/bbaf526 · Cited by 9
“Many algorithms require complete data matrices and thus various imputation techniques for handling missing data in bioinformatics have been explored in the context of multiomic imputation [98–100], including deep learning/AI for the integrative imputation of multiomic datasets [98].”
Gliozzo, J., Gomez, M. A. S., Bonometti, A., Patak, A., Casiraghi, E., & Valentini, G. (2025) · Bioinformatics · doi:10.1093/bioinformatics/btaf150 · Cited by 9
“…fusion methods are effective for disease subtype identification, biomarker discovery, and prognosis prediction, most cannot handle incomplete, or “partial”, datasets (Gliozzo et al. 2022, Flores et al. 2023) where one or more data sources are entirely missing for some patients (Supplementary Fig.”
Ramesh, D., Manickavel, P., Ghosh, S., & Bhat, M. (2025) · Journal of Clinical and Experimental Hepatology · doi:10.1016/j.jceh.2025.102611 · Cited by 9
Alwahaibi, N., & Alwahaibi, M. (2025) · Frontiers in Medicine · doi:10.3389/fmed.2025.1678753 · Cited by 8
“Reproducibility is hindered by variable sample preservation, sequencing depth, and analytical pipelines (186, 187).”
Sen, C. K., & DeMazumder, D. (2025) · Advances in Wound Care · doi:10.1177/21621918251380217 · Cited by 9
Integrating Machine Learning and Multi-Omics to Explore Neutrophil Heterogeneity
Lin, Z., Yang, T., Chen, D., Zhang, P., Luo, J., Chen, S., Gu, S., Shen, Y., Tang, T.-X., Chang, T., et al. (2025) · Biomedicines · doi:10.3390/biomedicines13092171 · Cited by 8
“Transparency and interpretability of AI and ML models are critical to promoting trust and accountability, as algorithms do not fully mimic real biological processes, which can erode clinician confidence and hinder clinical adoption [116].”
InfluentialSynOmics: integrating multi-omics data through feature interaction networks
Alif, M. N., & Zhang, W. (2025) · Briefings in Bioinformatics · doi:10.1093/bib/bbaf595 · Cited by 12
Thoniyot, S., & Balakrishnan, V. (2025) · Computational Biology and Chemistry · doi:10.1016/j.compbiolchem.2025.108772 · Cited by 9
Mtoor, S., Rashidian, N., Messaoudi, N., Grasso, V., Noël, F., Steindler, M., Jaradat, D., Frigerio, I. M., Butturini, G., Croner, R., et al. (2026) · Bioengineering · doi:10.3390/bioengineering13010117 · Cited by 7
“However, practical barriers remain substantial: multi-omics datasets differ in signal-to-noise ratios, suffer from missingness and partial observation, and often require imputation strategies that can introduce dependencies and bias if not carefully handled [72–74].”
EATRIS-Plus Multi-omics working group and Stakeholders, Alonso-Andrés, P., Baldazzi, D., Chen, Q., Conde, E., Crespo-Toro, L., Donner, K. M., Džubák, P., Ekberg, S., García‐Bermejo, L., et al. (2025) · Phenomics · doi:10.1007/s43657-024-00170-0 · Cited by 14
“While some integrative multi-omics methods can handle missing observations in some data modalities (Argelaguet et al. 2020; Cuklina et al. 2021), many methods require complete observations (Flores et al. 2023).”
How did we get there? AI applications to biological networks and sequences
Anteghini, M., Gualdi, F., & Oliva, B. (2025) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2025.110064 · Cited by 9
Machine Learning for Predicting Stroke Risk Stratification Using Multiomics Data: Systematic Review
Yoo, H. Y., Shin, H., Kim, E. E., & Son, Y.-J. (2026) · Journal of Medical Internet Research · doi:10.2196/85654 · Cited by 7
“Additionally, the definition of the analysis unit—for example, gene-level expression, CpG probe values, or LC-MS features—determined the imputation burden, multiplicity, and scope of feasible feature selection methods, thereby influencing both model performance and reproducibility [55,57,58].”
Multi-omics Quality Assessment in Personalized Medicine through EATRISPreprint
EATRIS Plus Multi-omics working group and stakeholders (in alphabetical order by last name), Alonso-Andrés, P., Baldazzi, D., Chen, Q., Conde, E., Crespo-Toro, L., Donner, K. M., Džubák, P., Ekberg, S., García‐Bermejo, L., et al. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.10.25.563912 · Cited by 6
Yang, T., & Li, W. V. (2025) · BMC Bioinformatics · doi:10.1186/s12859-025-06227-9 · Cited by 5
“Integrative methods for multi-modal data are typically categorized into three main approaches: early, middle, and late integration [7].”
Mukhopadhyay, S., Ulaganathan, N., Dumpuri, P., & Aich, P. (2025) · Methods in molecular biology · doi:10.1007/978-1-0716-4690-8_2 · Cited by 6
Genome-scale metabolic modelling in antimicrobial pharmacology: Present and future
Tao, J., Lin, Y.-W., Zhong, L., Zhu, Y., Yao, X., Aichem, M., Schreiber, F., Zhao, J., & Li, J. (2025) · Advanced Drug Delivery Reviews · doi:10.1016/j.addr.2025.115672 · Cited by 5
Zhong, H., Chi, S., Magaña, A. A., Fordwour, O. B., & Foster, L. J. (2025) · Journal of Proteome Research · doi:10.1021/acs.jproteome.5c00294 · Cited by 5
Mass Spectrometry Proteomics: A Key to Faster Drug Discovery
Tagliazucchi, L., & Costi, M. P. (2025) · Journal of Medicinal Chemistry · doi:10.1021/acs.jmedchem.5c01986 · Cited by 7
Liu, X., Bin, C., Gao, Q., Li, M., Xue, L., Zhou, Z., Liu, M., Ji, X., & Wei, S. (2026) · Biomarker Research · doi:10.1186/s40364-026-00914-2 · Cited by 8
“Strategy selection should, accordingly, prioritise approaches that tolerate missing blocks and avoid complete-case dependence [70, 454].”
A journey from omics to clinicomics in solid cancers: Success stories and challenges
Mehrotra, S., Sharma, S., & Pandey, R. K. (2024) · Advances in protein chemistry and structural biology · doi:10.1016/bs.apcsb.2023.11.008 · Cited by 6
Han, S., Yu, S., Shi, M., Harada, M., Ge, J., Lin, J.-S., Prehn, C. P., Petrera, A., Li, Y., Sam, F., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-58314-3 · Cited by 5
“The incompletenessof these datasets hindersmulti-omics integration 9 and investigations into predisposing factors (such as age and genetics), enabling factors (such as healthcare service and physical activity), and Fig.”
Zhou, W., Chen, Z., Liu, A., Zhang, X., Cao, X., Ding, Z., Sha, Q., Shen, H., & Deng, H. (2023) · Research Square · doi:10.21203/rs.3.rs-2768563/v1 · Cited by 5
Multi-omics and Artificial Intelligence technology for elucidating disease pathophysiology.Preprint
Jin, J., & Kim, J. J. (2024) · Preprints.org · doi:10.20944/preprints202408.0350.v1 · Cited by 5
Genetic and epigenetic drivers of neurodegenerative disorders
Roy, K., Syed, M., Narad, P., & Sengupta, A. (2025) · Progress in brain research · doi:10.1016/bs.pbr.2025.08.002 · Cited by 5
Okafor, U. C., Alghamdi, S. M., Anguilano, L., & Yang, Y. (2026) · Frontiers in Microbiology · doi:10.3389/fmicb.2026.1742848 · Cited by 9
“…from a lack of diversity in the soil-dwelling microbial taxa primarily due to the focus on model organisms that have been previously studied, and as a result, the microbial taxa that are present in soil that is tropical, arid, or chronically contaminated (Wang et al., 2023; Flores et al., 2023).”
A review of multi-omics integration techniques across five machine learning method families
Olowolayemo, A., Souag, A., Sirlantzis, K., Turner, S., & Wilson, C. M. (2026) · Bioinformatics Advances · doi:10.1093/bioadv/vbag108 · Cited by 5
“2022, Flores et al .”
InfluentialArtificial Intelligence-Based Methods and Omics for Mental Illness Diagnosis: A Review
de Oliveira, G. S., Rodrigues, F. H. D. S., Pontes, J. G. D. M., & Tasić, L. (2025) · Bioengineering · doi:10.3390/bioengineering12101039 · Cited by 4
“Beyond statistical analysis and model training, AI can also enhance different stages of omics workflows and facilitate multi-omics data integration [20].”
Multi-Omics Approaches for Reproductive Health Data
Menon, S. V., Joshi, G., Khadka, S., Joshi, P., Narayan, Y., Mahalingam, N. K., & Bhandari, M. (2024) · Springer eBooks · doi:10.1007/978-981-97-7451-7_4 · Cited by 4
Simplicity within biological complexity
Pržulj, N., & Malod‐Dognin, N. (2024) · Bioinformatics Advances · doi:10.1093/bioadv/vbae164 · Cited by 4
“…include: lack of robustness to noise and overfitting due to scarcity of labelled data, interpretability, and real networks usually being “multiplex” [201], i.e. having incomplete multi-view representations from different relations (some works recently attempting to address it [160, 161, 162, 202]).”
BackgroundChoudhury, P., Dasgupta, S., Bhattacharyya, P., Roychowdhury, S., & Chaudhury, K. (2024) · Molecular Omics · doi:10.1039/d3mo00266g · Cited by 3
Deep fusion of incomplete multi-omic data for molecular mechanism of Alzheimer’s disease
Xie, L., Raj, Y., Tong, M., Nho, K. T., Salama, P., Saykin, A. J., Fang, S., & Yan, J. (2025) · Scientific Reports · doi:10.1038/s41598-025-14636-2 · Cited by 4
Huang, W., Lin, R., Zeng, X., Wang, H., & Yan, J. (2025) · Frontiers in Immunology · doi:10.3389/fimmu.2025.1665688 · Cited by 4
Bioinformatics Strategies in Breast Cancer Research
Veneziano, M., Savini, I., Cortellesi, E., Gasperi, V., Gambacurta, A., & Catani, M. V. (2025) · Biomolecules · doi:10.3390/biom15101409 · Cited by 3
“…approaches (such as identifying relationships between biomarkers through models that better integrate well-known biological structures) and manages challenges like noisy data, missing information from some omics technologies, and reducing reliance on manual feature engineering [133,134].”
Mu, N., Yang, H., & Chen, Z. (2025) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2025.109179 · Cited by 7
Zivalj, D., Schleicher, L. M. S., Kršek, A., Diamee, H. J. F., Raljević, D., & Batičić, L. (2026) · Life · doi:10.3390/life16030400 · Cited by 4
Trends in the Application of Multiomics Based on Machine Learning in the Development of Probiotics
Chen, Y., Xing, Z., Sheng, J., Liu, X., Ding, W., & Chen, X. (2026) · Journal of Agricultural and Food Chemistry · doi:10.1021/acs.jafc.5c15387 · Cited by 5
Mapping the path to clinical implementation of multi-omics
Ismail, S. I., Saad, C., Elrayess, M. A., Khleifat, A. A., Hendi, N. N., Al-Sarraj, Y. A., Badji, R. M., & Al‐Muftah, W. (2026) · Nature Genetics · doi:10.1038/s41588-026-02663-2 · Cited by 5
“AI approaches, including deep learning and advanced imputation frameworks, are emerging as solutions for reconstructing missing information, integrating partially observed datasets and improving model robustness 57 .”
Decoding cell fate: integrated experimental and computational analysis at the single-cell level
Zhou, Y., Hou, S., Miao, X., Zhang, G., Li, Z., Zhang, D., Lin, Y., & Lin, Y. (2025) · Bioinformatics · doi:10.1093/bioinformatics/btaf603 · Cited by 4
“Building such context-aware, benchmarked frameworks will be essential for integrative analyses that remove technical artifacts while preserving biologically meaningful variation (Flores et al. 2023).”
ProteoArk: A One-Pot Proteomics Data Analysis and Visualization Tool for Biologists
Nisar, M., Soman, S. P., Sreelan, S., John, L., Pinto, S. M., Kandasamy, R. K., Subbannayya, Y., Prasad, T. S. K., Kanekar, S., Raju, R., et al. (2025) · Journal of Proteome Research · doi:10.1021/acs.jproteome.4c00556 · Cited by 3
Penalized likelihood optimization for censored missing value imputation in proteomics
Etourneau, L., Fancello, L., Wieczorek, S., Varoquaux, N., & Bürger, T. (2024) · Biostatistics · doi:10.1093/biostatistics/kxaf006 · Cited by 5
Liquid-based cytology in the era of multi-omics and artificial intelligence integration
Ahuja, S., & Zaheer, S. (2025) · CytoJournal · doi:10.25259/cytojournal_89_2025 · Cited by 2
“[9] In addition, the pre-analytical phase – including sample fixation, storage, and processing – can significantly influence molecular test performance.”
Li, D., Liao, K., Yan, B., Huang, J., Zhang, J., Chen, Y., Zhu, J., Zhi, S., & Chen, P. (2025) · Frontiers in Oncology · doi:10.3389/fonc.2025.1622600 · Cited by 6
Jin, D., & Saitō, Y. (2025) · Briefings in Bioinformatics · doi:10.1093/bib/bbaf698 · Cited by 2
“However, these datasets are often incomplete, with missing omics modalities in a substantial proportion of samples [8].”
Decoding disease and therapy through multiomics integration and systems analysis
Mathew, M. J., Mathew, J., Chacko, R. M., Bayry, J., & Zagury, J. (2026) · Briefings in Bioinformatics · doi:10.1093/bib/bbag311 · Cited by 5
“Despite advances, imputing missing values remains a challenge when the proportion of missing data is high and is varied across different omics datasets [ 82].”
Metadata
- Authors
- Flores, J. E., Claborne, D. M., Weller, Z. D., Webb-Robertson, B.-J. M., Waters, K. M., & Bramer, L. M.
- Affiliations
- Pacific Northwest National Laboratory (US)
- Corresponding author
- Bramer, Lisa M.
- Venue
- Frontiers in Artificial Intelligence
- Publisher
- Frontiers Media SA
- Published
- 2023-02-09
- Volume
- 6
- Pages
- 1098308-1098308
- ISSN
- 2624-8212 (ISSN-L)
- PMC
- PMC9949722
- OpenAlex
- W4319759388
- Semantic Scholar
- d833d42089ae664847832f8c0ef7bd7f6c1e26b6
- Cited by
- 181
- References
- 148
- Citations by year
- 2026: 55 · 2025: 77 · 2024: 38 · 2023: 11
- FWCI
- 4.78
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Frontiers in Artificial Intelligence (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
- Funders
- Pacific Northwest National Laboratory
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene expression and cancer classification
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
- Keywords
- missing data, multi-omics integration, artificial intelligence, statistical learning, missing data imputation, data heterogeneity, biological systems
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
