Protocol combining tree-based Maximum Parsimony and web-like Phylogenetic Networks analyses to investigate reticulate human evolution

Abstract

Our protocol combines Maximum Parsimony and Phylogenetic Networks approaches to understand the phylogenetic relationships and evolutionary processes of hominin species that might have shared inheritance from multiple ancestors. By addressing the questions of pattern and process in human phylogeny, the protocol can be used to clarify the taxonomic definition(s) of diverse hominin groups and ascertain whether or not the mode of evolution of genus Homo is reticulate. Using high quality and informative phenotypic data sets is necessary to yield meaningful results. For complete details on the use and execution of this protocol, please refer to Caparros and Prat (2021).

Keywords: Bioinformatics; Earth sciences; Evolutionary biology.

Graphical abstract

Figure 1

Protocol combining tree-based Maximum Parsimony and web-like Phylogenetic Networks analyses The three Major Steps of the protocol are as follows: (A) Tree-based MP analysis. (B) Intermediate tree-based MP analysis. (C) Reticulate PN analysis.

Figure 2

Character retention index ri formula and description of how it behaves as a function of changes in character state variables Yellow color: three variables m, s and g provided in the character diagnostic output of the first PAUP run. Pink color: homoplasy definitions based on variables m, s and g. Green color: behavior standardized from 0 to 1 of character retention index ri and its complement d observed homoplasy index. Blue color: inference interpretations drawn from behavior from 0 to 1 of character retention index ri. Reproduced from Figure S8 of Caparros and Prat (2021).

Figure 3

MPMAX cladogram and its elliptic representation MPMAX cladogram (A) and its elliptic representation (B) reproduced from Figures 1 and S7 of Caparros and Prat (2021). The nodes of the tree are indicated in black numbers with arrows. In the elliptic representation the LCA nodes are expressed as equivalent elliptic boundaries. Refer to the legend of Caparros and Prat (2021) Figure 1 for sister group definitions.

Figure 4

Character-state changes (apomorphies) in support of stem species elliptic boundaries 40 to 26 (tree nodes - LCAs) of MPMAX scenario Meaning of symbols are as follows: numbered stem boundaries correspond to nodes of equivalent tree, arrows from boxes to stem boundaries indicate supporting apomorphous character-state changes, ==> unambiguous characters and --> ACCTRAN resolved ambiguous characters. ACCTRAN optimization (Agnarsson and Miller, 2008) was used for the treatment of ambiguities. Reproduced from Caparros and Prat (2021) Figure S9.

Figure 5

Four examples of character state projections with MacClade based on the MPMAX cladogram Reproduced from Caparros and Prat (2021) Figure S8

Figure 6

Phylogenetic support comparison between MPMAX scenario and best BMCMC Dembo scenario (A) MPMAX bootstrap elliptic representation with replicate P_boot values (B) BMCMC Dembo elliptic representation with posterior probability values P_post. P_boot and P_post are shown in red and expressed in %. Reproduced from Caparros and Prat (2021) Figure 2.

Figure 7

50% majority consensus tree from MP run with apomorphous Data S2 that generated 3213 Dollo MP trees (MMPT) with RI=1 and CI=1 for all trees Reproduced from Caparros and Prat (2021) Figure 3.

Figure 8

Phylogenetic networks graphical results Tree and network computations were executed with the program SplitsTree (Huson and Bryant, 2006) based on Data S3 (MMPT = 3213 trees) obtained from the Intermediate MP analysis with Data S2. (A) Unrooted majority consensus tree with mean edge weights (pipeline Taxa=24 [Trees > ConsensusTree > EqualAngle]). (B) Consensus network computed with threshold = 0.10, mean edge weights and splits convex hull transformation (pipeline Taxa=24 [Trees > ConsensusNetwork > ConvexHull]). (C) Rooted reticulate network with first taxon as outgroup and 180 equal angle (pipeline Taxa=24 [Trees > ConsensusNetwork > ReticulateNetwork]). Reproduced from Figure 4 in Caparros and Prat (2021).

Figure 9

Highlight of multiple trees conflicting patterns expressed as polytomies in consensus tree, and expressed in the consensus network by means of split parallelograms that may induce reticulation Reproduced from Caparros and Prat (2021) Figure 4.

Figure 10

Inferences drawn with regard to the modes of evolution of the group of hominin species covering the period Late Miocene to the Holocene Reproduced from Caparros and Prat (2021) Figure 4.