The apps

cogent3 comes with pre-defined “apps” which perform single functions that can be configured by the user. Apps are designed to enable usage without requiring detailed understanding of programming and so they can be applied in batch to a large amount of data. In other words, being able to use them as functions without writing loops or conditionals.

Apps can be used by themselves, or added together to define a “composed function” (aka pipeline). Composed functions can be applied to a single, or thousands, of data file(s).

Apps have several key features

1. trap errors

2. check the validity of input data

3. can be used by themselves or combined into a “composed” app

4. automatically track the relationship between an input data record and its output record

Overview

• Overview of using apps
• Getting help on an app
• Getting an app
• Displaying installed apps
• Specifying data for analysis
• Tracking records that could not be processed

The progressive_align App

cogent3 has a built-in progressive aligner. This is a tree-based HMM aligner that can use any cogent3 substitution model.

• Using a nucleotide model
• Using a codon model
• Using a protein model

The model App

cogent3 possesses unique capabilities for molecular evolutionary analyses. Of particular significance are our non-stationary model classes and the associated measurement of genetic distance from these. That said, we also implement most of the conventional substitution models. See cogent3.available_models() for a display of the built-in models.

You can use these models either through the general model app, or some task specific apps. Both are described below.

• Apply a non-stationary nucleotide model to an alignment with 3 sequences
• Apply a non-stationary nucleotide model to an alignment with a tree
• Specifying a non-stationary model with time-heterogeneous parameters
• Extracting maximum likelihood estimates from a model_result
• Reconstructing ancestral states
• Applying a discrete-time, non-stationary nucleotide model
• Applying a time-reversible nucleotide model
• Applying GNC, a non-stationary codon model
• Applying a time-reversible codon model

The hypothesis App

Takes two (or more) model instances and performs hypothesis test.

• Testing a hypothesis – non-stationary or time-reversible

The natsel Apps

These apps all employ codon models to evaluate different types of hypotheses regarding the mode of natural selection.

• natsel_neutral – a test for selective neutrality
• natsel_timehet – a test of branch heterogeneity
• natsel_sitehet – a test of site heterogeneity
• natsel_zhang – a branch-site test

Write your own Apps

You can write your own apps.

• Turn your functions into composable apps