PROTTEST(1)

💖 PROTTEST(1) — the girlies' guide to picking your best-fit protein evolution model 💖

okay so bestie, welcome to prottest — aka the sequential version of "Selection of best-fit models of protein evolution". Yes she's a whole manual page, yes she's from User Commands, and yes she is about to change your whole phylogenetics situation. 🧬✨

🌸 NAME 🌸

prottest — Selection of best-fit models of protein evolution (sequential version). that's it, that's the girl. 💅

✨ SYNOPSIS ✨

prottest -i alignm_file [OPTIONS]

literally just give her an alignment and some options and she goes. periodt.

🦋 DESCRIPTION 🦋

PROTTEST is ModelTest's relative — no cap — and what she does is pick the model of protein evolution that best fits a given set of sequences (an alignment). she's a java program, and she's built on the Phyml program for maximum likelihood calculations and optimization of parameters, plus she uses the PAL library too. iconic lineage honestly.

the models she plays with are:

• empirical substitution matrices — these tell you the relative rates of amino acid replacement: WAG, LG, mtREV, Dayhoff, DCMut, JTT, VT, Blosum62, CpREV, RtREV, MtMam, MtArt, HIVb, and HIVw
• specific improvements — to account for the evolutionary constraints impossed by conservation of protein structure and function: +I (invariable sites), +G (rate heterogeneity among sites), and +F (observed amino acid frequencies)

then she uses the Akaike Information Criterion (AIC) and other statistics (AICc and BIC) to find which of the candidate models best fits the data at hand. she's basically a matchmaker but for models and data. 🫶

👑 OPTIONS 👑

• -i alignment_filename — the alignment input file. required, so like. don't forget her. 😭
• -t tree_filename — tree file. optional! [default: NJ tree]
• -o output_filename — output file. optional! [default: standard output]
• -log enabled/disabled — enables / disables PhyML logging into the log directory (see prottest.properties)
• -[matrix] — include matrix (Amino-acid) = JTT LG DCMut MtREV MtMam MtArt Dayhoff WAG RtREV CpREV Blosum62 VT HIVb HIVw FLU

if you don't specify any matrix, all matrices displayed above will be included. she's giving generous 💅

• -I — include models with a proportion of invariable sites
• -G — include models with rate variation among sites and number of categories
• -IG — include models with both +I and +G properties
• -all-distributions — include models with rate variation among sites, number of categories and both
• -ncat number_of_categories — define number of categories for +G and +I+G models [default: 4]
• -F — include models with empirical frequency estimation
• -AIC — display models sorted by Akaike Information Criterion (AIC)
• -BIC — display models sorted by Bayesian Information Criterion (BIC)
• -AICC — display models sorted by Corrected Akaike Information Criterion (AICc)
• -DT — display models sorted by Decision Theory Criterion
• -all — displays a 7-framework comparison table
• -S optimization_strategy — optimization strategy mode: [default: 0]
  • 0: Fixed BIONJ JTT
  • 1: BIONJ Tree
  • 2: Maximum Likelihood tree
  • 3: User defined topology
• -s moves — tree search operation for ML search: NNI (fastest), SPR (slowest), BEST (best of NNI and SPR) [default: NNI]
• -t1 — display best-model's newick tree [default: false]
• -t2 — display best-model's ASCII tree [default: false]
• -tc consensus_threshold — display consensus tree with the specified threshold, between 0.5 and 1.0 [0.5 = majority rule consensus ; 1.0 = strict consensus]
• -threads number_or_threads — number of threads requested to compute (only if MPJ is not used) [default: 1]
• -verbose — verbose mode [default: false]

🎀 EXAMPLE 🎀

prottest -i alignm_file -t tree_file -S 0 -all-distributions -F -AIC -BIC -tc 0.5 > output

see how she ate that? alignment in, tree in, strategy 0, all the distributions, frequencies, AIC, BIC, consensus threshold 0.5, and everything redirected into output. main character energy fr fr. 🌈

and that's the whole vibe bestie — this was prottest 3.4, brought to you in October 2015 🎂 (yes she's an elder but she's still slaying). go run your alignment and let her pick your model. you got this 💜