https://github.com/cmdoret/strangerseq

Generate DNA sequences minimizing microhomology
https://github.com/cmdoret/strangerseq

dna-sequences genomics k-mer

Last synced: 6 days ago
JSON representation

Generate DNA sequences minimizing microhomology

• Host: GitHub
• URL: https://github.com/cmdoret/strangerseq
• Owner: cmdoret
• License: gpl-3.0
• Created: 2019-07-11T12:28:25.000Z (almost 6 years ago)
• Default Branch: master
• Last Pushed: 2021-03-05T13:46:09.000Z (about 4 years ago)
• Last Synced: 2025-02-17T23:47:36.022Z (3 months ago)
• Topics: dna-sequences, genomics, k-mer
• Language: Go
• Size: 48.8 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
### strangerseq

[![CircleCI](https://circleci.com/gh/cmdoret/strangerseq/tree/master.svg?style=svg)](https://circleci.com/gh/cmdoret/strangerseq/tree/master)

**cmdoret**

strangerseq is a Go program that generates DNA sequences minimizing or maximizing microhomology to a reference genome.

This is achieved using an l-order Markov chain from the K-mer profile of the genome (l = k-1). Sequences are initiated

by picking a random k-mer, with their frequency used as probability weight. Extension is then performed iteratively using

the markov chain. When minimizing microhomology, inverse probabilities from the chain are used to pick rare k-mers instead

of frequent ones.

It is possible to add a GC deviation constraint to the sequences to force GC content to be similar to the genome.

#### Usage

```

strangerseq -help

Program to generate sequences with minimal microhomology and optionally, similar GC content to the input genome.

Multiple sequences are generated and sorted by score.

  -comp.seq

        Enable to return scores in addition to sequences and include randomly generated GC-weighted sequences for comparison. Columns of the output are: 1. sequence type (generated through markov model or randomly picked with GC weight), 2. Score without accounting for GC divergence, 3. Score corrected for GC divergence, 4. Sequence.

  -fasta string

        Path to genome file in FASTA format. (required)

  -fixed.gc float

        Fixed target GC content to use as target. The default is to use the input genome's GC content.

  -gc.weight float

        Weight given to the GC content when scoring sequences. (default 1)

  -kmer.size int

        Length of K-mers on which to optimize sequences. (default 8)

  -n.seq int

        Number of sequences to generate. (default 100)

  -seq.len int

        Length of the sequences to generate. (default 1000)

  -similar

        Generate similar sequences (frequent k-mers) instead of different ones (rare k-mers).

  -version

        Shows version number of the binary.

```

#### Example

Generate 50 GC-constrained sequences of 1000 base pairs minimizing microhomology:

```bash

./strangerseq -fasta genome.fa -gc.weight 1 -n.seq 50

```