Editing Molecular phylogenetics (section)

==Limitations==
Molecular systematics is an essentially [[Cladistics|cladistic]] approach: it assumes that classification must correspond to phylogenetic descent, and that all valid taxa must be [[monophyletic]]. This is a limitation when attempting to determine the optimal tree(s), which often involves bisecting and reconnecting portions of the phylogenetic tree(s).

The recent discovery of extensive [[horizontal gene transfer]] among organisms provides a significant complication to molecular systematics, indicating that different genes within the same organism can have different phylogenies. HGTs can be detected and excluded using a number of phylogenetic methods (see {{section link|Inferring horizontal gene transfer|Explicit phylogenetic methods}}).

In addition, molecular phylogenies are sensitive to the assumptions and models that go into making them. Firstly, sequences must be aligned; then, issues such as [[long-branch attraction]], [[saturation (genetic)|saturation]], and [[taxon]] sampling problems must be addressed. This means that strikingly different results can be obtained by applying different models to the same dataset.<ref>{{cite journal | doi = 10.1093/zoolinnean/zlz088| title = Exploring the impact of morphology, multiple sequence alignment and choice of optimality criteria in phylogenetic inference: A case study with the Neotropical orb-weaving spider genus Wagneriana (Araneae: Araneidae)| journal = Zoological Journal of the Linnean Society| year = 2020| volume = 188 | issue = 4 | pages = 976–1151 | last1 = Cabra-García| first1 = Jimmy| last2 = Hormiga| first2 = Gustavo}}</ref><ref name="Philippe2011">{{Cite journal |last1=Philippe |first1=H. |last2=Brinkmann |first2=H. |last3=Lavrov |first3=D. V. |last4=Littlewood |first4=D. T. J. |last5=Manuel |first5=M. |last6=Wörheide |first6=G. |last7=Baurain |first7=D. |editor1-last=Penny |editor1-first=David |title=Resolving Difficult Phylogenetic Questions: Why More Sequences Are Not Enough |doi=10.1371/journal.pbio.1000602 |journal=PLOS Biology |volume=9 |issue=3 |pages=e1000602 |year=2011 |pmid=21423652 |pmc=3057953 |doi-access=free }}</ref> The tree-building method also brings with it specific assumptions about tree topology, evolution speeds, and sampling. The simplistic UPGMA assumes a rooted tree and a uniform molecular clock, both of which can be incorrect.<ref name="Pevsner" />

The low resolution power of single genes have been overcome using [[Computational phylogenetics#Multigene phylogeny|multigene phylogenies]], though the issue of HGT still merits careful algorithmic design.<ref>{{cite web |title=Part 4: Multigene phylogenetics |url=https://web.natur.cuni.cz/~vlada/moltax/part4.html |website=web.natur.cuni.cz |language=en}}</ref>