HYPHY(1)

🌸 HYPHY(1) 🌸

✨ NAME ✨

okay so the whole deal is hyphy 💖, a tool for hypothesis testing using phylogenies, and this is the pthreads version 🧚

✨ DESCRIPTION ✨

bestie, this is the usage line 👑

usage: hyphy or HYPHYMPI [-h] [--help][-c] [-d] [-i] [-p]
       [BASEPATH=directory path] [CPU=integer] [LIBPATH=library path]
       [USEPATH=library path] [<standard analysis name> or <path to hyphy batch
       file>] [--keyword value ...] [positional arguments ...]

So you can run a HyPhy analysis either interactively or in batch mode 🦋

🎀 optional flags

• -h, --help — shows the help message and exits 🙈
• -c — calculator mode! HyPhy drops you into expression evaluation until you type exit 🧮
• -d — debug mode 🐛 HyPhy drops you into expression evaluation mode if a script errors
• -i — interactive mode ✨ HyPhy always asks you for analysis options, even when defaults are available
• -p — postprocessor mode 🌈 you get an interactive mode where general post-processing scripts can be picked once the analysis finishes
• -m — writes diagnostic messages to messages.log 📝

🌈 optional global arguments

• BASEPATH=directory path — the base directory for all path operations (default is pwd) 📂
• CPU=integer — if it was compiled with OpenMP multithreading support, this asks for that many threads 🧵. HyPhy could use fewer than you ask for, but never more. The default is the number of CPU cores (as computed by OpenMP) on your system 💪
• LIBPATH=directory path — where the HyPhy library files live (default installed location is /usr/local/lib/hyphy or whatever was configured during CMake installation) 📚
• USEPATH=directory path — the optional working and relative path directory (default is BASEPATH) 🗺️
• ENV=expression — sets HBL environment variables with explicit statements, like ENV='DEBUG_MESSAGES=1;WRITE_LOGS=1' 🛠️
• batch file to run — if you specify one, HyPhy runs that file. Otherwise it drops you into interactive mode 🎬
• analysis arguments — if a batch file is present, every remaining positional argument gets read as an input to the analysis prompts 🧃

🫶 optional keyword arguments

These can show up anywhere in the command, and the requested analysis is the one that eats them 🍽️

• --keyword value — gets passed to the analysis (the one that uses KeywordArgument directives). If you use the same keyword multiple times, those values become an array of values for multiple selectors 🎯

🍓 usage examples

Select a standard analysis from the list:

hyphy -i

Run a standard analysis with default options and one required user argument:

hyphy busted --alignment path/to/file

Run a standard analysis with additional keyword arguments:

hyphy busted --alignment path/to/file --srv No

See which arguments a standard analysis understands:

hyphy busted --help

Run a custom analysis and pass it some arguments:

hyphy path/to/hyphy.script argument1 'argument 2'

👑 Available standard keyword analyses

These live in /home/nilesh/packages/hyphy/hyphy/res/, and here's the lineup 💅

• meme — [MEME] tests for episodic site-level selection using MEME (Mixed Effects Model of Evolution) 🔥
• mh — merges two datafiles by combining sites (horizontal merge) 🧩
• mv — merges two datafiles by combining sequences (vertical merge) 🧩
• mcc — compares mean within-clade branch length or pairwise divergence between two or more non-nested clades in a tree 🌳
• mclk — tests for the presence of a global molecular clock on the tree using its root. The resulting clock tree is unrooted, but one of the root branches can be divided so the clock gets enforced ⏰
• mgvsgy — compares the fits of MG94 and GY94 models (crossed with an arbitrary nucleotide bias) on codon data 🧬
• mt — selects an evolutionary model for nucleotide data using the methods of 'ModelTest', a program by David Posada and Keith Crandall 🧪
• fel — [FEL] tests for pervasive site-level selection using FEL (Fixed Effects Likelihood) 💖
• fubar — [FUBAR] tests for pervasive site-level selection using FUBAR (Fast Unconstrained Bayesian AppRoximation for inferring selection) 🦋
• fade — [FADE] tests a protein alignment for directional selection towards specific amino acids along a set of test branches you specify, using FADE (a FUBAR Approach to Directional Evolution) 🎯
• faa — fits a multiple fitness class model to amino acid data 🧫
• fmm — "Fit a model that permits double (and triple) instantaneous nucleotide substitutions" 🌀
• fst — computes various measures of F_ST and (optionally) performs permutation tests 🎲
• slac — [SLAC] tests for pervasive site-level selection using SLAC (Single Likelihood Ancestor Counting) 🔍
• sm — performs a classic and structured Slatkin-Maddison test for the number of migrations 🦩
• sns — parses a codon alignment for ambiguous codons and outputs a complete list/resolutions/syn and ns counts by sequence/position 📋
• sw — performs a sliding window analysis of sequence data 🪟
• sa — performs a phylogeny reconstruction for nucleotide, protein or codon data with user-selectable models, using the method of sequential addition 🌱
• sbl — searches an alignment for a single breakpoint 💔
• spl — plots genetic distances (similarity) of one sequence against all the others in an alignment, using a sliding window. Optionally, it figures out NJ-based clustering and bootstrap support in every window. This is a HyPhy adaptation of the excellent (but Windows-only) SimPlot (and/or VarPlot) by Stuart Ray (http://sray.med.som.jhmi.edu/SCRoftware/simplot/) 📈
• busted — [BUSTED] tests for episodic gene-wide selection using BUSTED (Branch-site Unrestricted Statistical Test of Episodic Diversification) 💅
• bgm — [BGM] applies Bayesian Graphical Model inference to substitution histories at individual sites 🕸️
• bva — runs a selection analysis using a general discrete bivariate (dN AND dS) distribution. The right number of rate classes gets figured out automatically 🤖
• brp — interprets bivariate codon rate analysis results 📊
• bsel — splits a tree into two clades (compartments) plus a separating branch, then tests for equal dN/dS between compartments and for selection along the separating branch, using a series of Likelihood Ratio Tests ✂️
• bst — uses the improved branch-site REL method of Yang and Nielsen (2005) to look for episodic selection in sequences 🔬
• bt — tests whether a branch (or branches) in the tree evolves under different dN and dS than the rest of the tree 🌿
• absrel — [aBSREL] tests for lineage-specific evolution using the branch-site method aBS-REL (Adaptive Branch-Site Random Effects Likelihood) 🌟
• acd — analyses codon data with a variety of standard models using a given tree 🧬
• ad — analyses nucleotide or aminoacid data with a variety of standard models using a given tree 🧪
• adn — analyses di-nucleotide data with a variety of standard models using a given tree 🫧
• afd — analyses nucleotide data with a variety of standard models using a given tree, estimating equilibrium frequencies as parameters ⚖️
• ana — runs a selection analysis 🎀
• ai — Peter Simmonds' Association Index (AI) 🧠
• relax — [RELAX] tests for relaxation of selection pressure along a specified set of test branches using RELAX (a random effects test of selection relaxation) 🦋
• clst — clustering time! 🌸 Apply clustering methods for phylogeny reconstruction (UPGMA, WPGMA, complete or minimal linkage) to nucleotide, protein and codon data. It uses MLE of pairwise distances with models you pick yourself, and the trees it makes come with a global molecular clock. ⏰
• leisr — infer relative evolutionary rates 🦋 on a nucleotide or protein alignment, kinda like Rate4Site (PMID: 12169533). 💅
• lz — Lempel-Ziv complexity and entropy 🧚 of sequences, even if they aren't aligned. 🌈
• lclk — test for a local molecular clock ⭐. Every subtree of your tree gets the clock constraint, while the rest of the tree stays clock-free. It's giving selective clock vibes 👑
• lht — a likelihood ratio test 💜 to detect conflicting phylogenetic signal (Huelsenbeck and Bull, 1996). [Contributed by Olivier Fedrigo] 🫶
• tc — check whether a group of sequences in a sample clusters together 🎀
• ts — do an exhaustive tree space search for nucleotide or protein data with models you choose. 🔥

⚠️ bestie, this one is only for data sets with less than 10 taxa! Anything bigger and it's a no from me 💀

• dtr — read sequence data (#, PHYLIP, NEXUS) and convert it to a different format 🍓
• dist — make a pairwise sequence distance matrix in PHYLIP format ✨
• kh — run a Kishino-Hasegawa test on two competing phylogenies 🥊
• ub — get an upper bound on the likelihood score of a dataset 🌸
• nuc — compare all 203 reversible nucleotide models and run LRT and AIC model selection. Main character energy 👑
• nj — phylogeny reconstruction for nucleotide, protein or codon data with user-selectable models, using neighbor joining. 🌷
• ny — test for positive selection with the approach of Nielsen and Yabg 🧪. It samples global dN/dS from an array of distributions and uses a Bayesian posterior to find the sites with dN/dS>1. Iykyk 🙈
• gard — [GARD] screen an alignment using GARD. You need an MPI environment for this one 🌈

---

✨ AUTHOR ✨

This manpage was written by Nilesh Patra for the Debian distribution and can be used for any other usage of the program. 🫶

---

And that's the whole page, bestie! 💖 This one's from hyphy 2.5.26, dated January 2021, and it's HYPHY(1) all the way 🦋🌸