The common ancestry of life

“We then split each alignment into two alignments of 10 sequences each and reconstructed phylogenetic trees for each of the resulting 300 alignments (100 original ones plus two halves for each; Figure 1) using the PhyML program (WAG evolutionary model, alpha parameter of gamma distribution set to 1) [10].”

“First, the available experimental studies, however limited, suggest that, although only a small fraction of the vast sequence space is open for evolution, the available trajectories are nevertheless numerous, so that evolution is far from deterministic [12,13].”

“However, formal demonstration of UCA, independent of the assumption that universally conserved orthologous proteins with highly similar sequences actually originate from common ancestral forms, remains elusive and might not be feasible in principle.”

“Second, the few described cases of actual convergent evolution of similar protein sequences, resulting in independent emergence of the same enzyme specificity, involve only a few key amino acid residues and do not attest to convergent origin of highly similar sequences [14,15].”

“Doolittle [11], where this is laid out in more detail.”

“However, formal demonstration of UCA, independent of the assumption that universally conserved orthologous proteins with highly similar sequences actually originate from common ancestral forms, remains elusive and might not be feasible in principle.”

“The likelihood tests of the kind described by Theobald [4] fail to address the problem because they yield results “in support of common ancestry” for any sufficiently similar sequences.”

“Second, the few described cases of actual convergent evolution of similar protein sequences, resulting in independent emergence of the same enzyme specificity, involve only a few key amino acid residues and do not attest to convergent origin of highly similar sequences [14,15].”

“First, the available experimental studies, however limited, suggest that, although only a small fraction of the vast sequence space is open for evolution, the available trajectories are nevertheless numerous, so that evolution is far from deterministic [12, 13 ].”

“Third, and perhaps most convincing, for about 10% of the known enzyme specificities, isofunctional proteins without detectable sequence or structure similarity have been detected, an indication that multiple, independent solutions to the same biological function are accessible to evolution [16,17].”

“Response: We agree and incorporated this point into the revised version of the article: “Following the general information theoretical framework for statistical tests of common ancestry laid out previously by Sober and Steel [6], Theobald reports a likelihood ratio test of the common ancestry hypothesis for genes repr…”

“In a recent, remarkable Letter to Nature, Theobald applied an information-theoretical approach to offer just that: a formal, homology-independent test for the hypothesis of the common ancestry of the extant cellular life forms [4], a claim that is further reaffirmed in the accompanying News and Views article by Steel…”

“Recent analyses based on phylogenetic model selection of universal common ancestry versus independent origin scenarios have supported universal common ancestry (Theobald 2010, 2011), although these findings are debated (Koonin and Wolf 2010; Yonezawa and Hasegawa 2010; Martins and Posada 2012).”

“It cannot be excluded that protein similarities revealed by using BLAST only cannot provide a basis for phylogenetic inferences [55, 56].”

“PGK can be traced back to LUCA [1,2,16] and its evolutionary rate is extremely slow.”

“Claim 2: Koonin and Wolf’s simulation corresponds to a common ancestry model K&W assert, without proof, that the profile model used in their simulation produces “phylogenetically unrelated sequences”.”

“This test spawned several criticisms and followon discussions (Koonin and Wolf 2010; Steel and Penny 2010; Theobald 2010b, 2011; Yonezawa and Hasegawa 2010, 2012; Martins and Posada 2014).”

“Koonin and Wolf have recently exposed the fallacy of taking similarity as proof of homology [64], and yet in our judgment they commit another fallacy.”

“However, likelihood analyses may be illusory in proving the hypothesis [9], and at least ambiguous in constructing one tree of life [10].”

“Within most neutral models of evolution, the probability that a sequence appears in two unrelated genomes decays exponentially with its length, so that sufficiently long sequences common to more than one genome are expected to derive from a common ancestor [15].”

“Koonin and Wolf (2010) simulated alignments lacking phylogenetic structure (site columns came from an independent distribution of amino acid frequencies) and showed that the test would spuriously favor UCA, probably because it was misled by column-wise similarity.”

“…ancestor with a formal quantitative testing (Theobald, 2010; Penny et al., 1982), although an agreement about if this can formally be tested has not been achieved (Velasco, 2018; de Oliveira Martins & Posada, 2016; Koskela & Annila, 2012; Koonin & Wolf, 2010; Theobald, 2010; Penny et al., 1982).”

“Given such a loss of information, the codon usage test is not useful for testing the universal common ancestry (UCA) of all life on earth, which is well known asa difficult problem (Theobald 2010; Koonin and Wolf 2010; Yonezawa and Hasegawa 2010, 2012; de Oliveira Martins and Posada 2014).”