README.md 10.3 KB
Newer Older
RomainFeron's avatar
RomainFeron committed
1
2
[![GitHub release (latest by date)](https://img.shields.io/github/v/release/SexGenomicsToolkit/RADSex?color=lightorange)](https://github.com/SexGenomicsToolkit/RADSex/releases)
[![Conda (channel only)](https://img.shields.io/conda/vn/bioconda/radsex?color=lightorange)](https://bioconda.github.io/recipes/radsex/README.html)
RomainFeron's avatar
RomainFeron committed
3
[![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.4628297.svg)](https://doi.org/10.5281/zenodo.4628297)
4

RomainFeron's avatar
RomainFeron committed
5
# radsex
6

Romain Feron's avatar
Romain Feron committed
7
## Overview
8

RomainFeron's avatar
RomainFeron committed
9
10
11
The `radsex` software is part of RADSex, a computational workflow for the analysis of sex-determination using RAD-Sequencing data. This workflow contains the software `radsex` and the R package `sgtr`; a Snakemake implementation of the workflow is available [here](https://github.com/SexGenomicsToolkit/RADSex-workflow).

The first step of the RADSex workflow is to use `radsex` to generate a summary file for a set demultiplexed RAD reads and use this file to:
RomainFeron's avatar
RomainFeron committed
12
13
14
15
16
17

- infer the type of sex-determination system
- identify sex-biased markers
- align markers to a genome and identify genomic regions differentiated between sexes
- compute marker depth statistics

RomainFeron's avatar
RomainFeron committed
18
Results from `radsex` can be visualized with the `sgtr` R package available [here](https://github.com/SexGenomicsToolkit/sgtr).
RomainFeron's avatar
RomainFeron committed
19

RomainFeron's avatar
RomainFeron committed
20
Although RADSex was developed specifically to study sex-determination, it was designed to be flexible and can be used to compare two groups for any binary trait.
21

RomainFeron's avatar
RomainFeron committed
22
The RADSex computational workflow was developed in the [LPGP](https://www6.rennes.inra.fr/lpgp/) lab from INRA, Rennes, France for the PhyloSex project, which investigates sex determining factors in a wide range of fish species.
Romain Feron's avatar
Romain Feron committed
23

24
25
### Citing RADSex

RomainFeron's avatar
RomainFeron committed
26
If you use RADSex in your work, please cite the [manuscript describing RADSex](https://onlinelibrary.wiley.com/doi/abs/10.1111/1755-0998.13360):
27

RomainFeron's avatar
RomainFeron committed
28
> Feron, R., Pan, Q., Wen, M., Imarazene, B., Jouanno, E., Anderson, J., Herpin, A., *et al.* (2021), RADSex: A computational workflow to study sex determination using restriction site‐associated DNA sequencing data. Mol Ecol Resour. https://doi.org/10.1111/1755-0998.13360
29
30
31
32
33
34


To properly cite RADSex, you should also cite the software and its version using the DOI provided in the badges above:

> Romain Feron. (2020, April 29). SexGenomicsToolkit/radsex: 1.1.2 (Version 1.1.2). Zenodo. http://doi.org/10.5281/zenodo.3775206

35
## Documentation
Romain Feron's avatar
Romain Feron committed
36

RomainFeron's avatar
RomainFeron committed
37
This README file includes a basic installation guide and a quick start section. The full documentation for RADSex, including a complete example walkthrough, is available [here](https://sexgenomicstoolkit.github.io/html/radsex/introduction.html).
Romain Feron's avatar
Romain Feron committed
38

39
40
Internal documentation generated with Doxygen is provided [here](https://sexgenomicstoolkit.github.io/radsex).

RomainFeron's avatar
RomainFeron committed
41
## Installation
Romain Feron's avatar
Romain Feron committed
42

RomainFeron's avatar
RomainFeron committed
43
### Requirements
44

RomainFeron's avatar
RomainFeron committed
45
46
RADSex was tested on Linux (ubuntu >= 18.04, Arch) and OSX. To install RADSex, you will need:

47
- A C++11 compliant compiler (GCC >= 6.1.0, Clang >= 6.0)
RomainFeron's avatar
RomainFeron committed
48
- The zlib library (usually installed on linux and osx by default)
Romain Feron's avatar
Romain Feron committed
49

RomainFeron's avatar
RomainFeron committed
50
### Install the latest official release
51

RomainFeron's avatar
RomainFeron committed
52
- Download the [latest release](https://github.com/SexGenomicsToolkit/radsex/releases)
RomainFeron's avatar
RomainFeron committed
53
- Unzip the archive
RomainFeron's avatar
RomainFeron committed
54
- Navigate to the `radsex` directory
55
- Run `make`
RomainFeron's avatar
RomainFeron committed
56

RomainFeron's avatar
RomainFeron committed
57
The compiled `radsex` binary will be located in **radsex/bin/**.
RomainFeron's avatar
RomainFeron committed
58
59
60
61

### Install the latest stable development version

```bash
RomainFeron's avatar
RomainFeron committed
62
63
git clone https://github.com/SexGenomicsToolkit/radsex.git
cd radsex
RomainFeron's avatar
RomainFeron committed
64
65
66
make
```

RomainFeron's avatar
RomainFeron committed
67
The compiled `radsex` binary will be located in **radsex/bin/**.
RomainFeron's avatar
RomainFeron committed
68

RomainFeron's avatar
RomainFeron committed
69
### Install radsex with Conda
RomainFeron's avatar
RomainFeron committed
70

RomainFeron's avatar
RomainFeron committed
71
All versions of radsex are available in [Bioconda](https://bioconda.github.io/recipes/radsex/README.html?#recipe-Recipe%20'radsex'). To install the latest radsex release with Conda, run the following command:
RomainFeron's avatar
RomainFeron committed
72
73
74
75

```bash
conda install -c bioconda radsex
```
76

77
78
## Quick start

RomainFeron's avatar
RomainFeron committed
79
### Preparing the data
80

RomainFeron's avatar
RomainFeron committed
81
Before running the workflow, you should prepare the following elements:
RomainFeron's avatar
RomainFeron committed
82

83
84
85
- A **set of demultiplexed reads**. The current version of RADSex does not implement demultiplexing;
  raw sequencing reads can be demultiplexed using [Stacks](http://catchenlab.life.illinois.edu/stacks/comp/process_radtags.php)
  or [pyRAD](http://nbviewer.jupyter.org/gist/dereneaton/af9548ea0e94bff99aa0/pyRAD_v.3.0.ipynb#The-seven-steps-described).
86

RomainFeron's avatar
RomainFeron committed
87
- A **group info file** (popmap): a tabulated file with individual IDs in the first column and sex (or group) in the second column. Individual IDs in the popmap must be the same as the names of the demultiplexed reads files (*e.g.* 'individual_1' for the reads file 'individual_1.fq.gz')
RomainFeron's avatar
RomainFeron committed
88
89

- To align the markers to a genome: the **genome** sequence in a FASTA file.
RomainFeron's avatar
RomainFeron committed
90
  Note that when visualizing `map` results with `sgtr`, linkage groups / chromosomes are automatically inferred from scaffold names in the genome if their name starts with *LG*, *CHR*, or *NC* (case unsensitive). If chromosomes are named differently in the genome, you can use a tabulated file with contig ID in the first column and corresponding chromosome name in the second column (see the doc for details).
RomainFeron's avatar
RomainFeron committed
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106

### Computing the marker depths table

The first step of RADSex is to create a table of marker depths for the entire dataset using the `process` command:

```bash
radsex process --input-dir ./samples --output-file markers_table.tsv --threads 16 --min-depth 1
```

In this example, demultiplexed reads are located in **./samples** and the markers table generated by `process` will be saved to **markers_table.tsv**. The parameter `--threads` specifies the number of threads to use, and `--min-depth` specifies the minimum depth to consider a marker present in an individual: markers which are not present with depth higher than this value in at least one individual will not be retained in the markers table.
It is advised to keep the minimum depth to the default value of 1 for this step, as it can be adjusted for each analysis later.


### Computing the distribution of markers between sexes

The `distrib` command computes the distribution of markers between males and females from a marker depths table:
107

RomainFeron's avatar
RomainFeron committed
108
109
110
```bash
radsex distrib --markers-table markers_table.tsv --output-file distribution.tsv --popmap popmap.tsv --min-depth 5 --groups M,F
```
111

RomainFeron's avatar
RomainFeron committed
112
In this example, `--markers-table` is the table generated with `process` and the distribution of markers between males and females will be saved to **distribution.tsv**. The sex of each individual in the population is given by **popmap.tsv**. Groups of individuals to compare (as defined in the popmap) are specified manually with the parameter `--groups`. The minimum depth to consider a marker present in an individual is set to 5, meaning that markers with depth lower than 5 in an individual will not be considered present in this individual.
113

RomainFeron's avatar
RomainFeron committed
114
The resulting distribution can be visualized with the `radsex_distrib()` function of [sgtr](https://github.com/SexGenomicsToolkit/sgtr), which generates a tile plot of marker counts with number of males on the x-axis and number of females on the y-axis.
115

RomainFeron's avatar
RomainFeron committed
116
### Extracting markers significantly associated with sex
117

RomainFeron's avatar
RomainFeron committed
118
Markers significantly associated with sex are obtained with the `signif` command:
119

RomainFeron's avatar
RomainFeron committed
120
121
122
```bash
radsex signif --markers-table markers_table.tsv --output-file markers.tsv --popmap popmap.tsv --min-depth 5 --groups M,F [ --output-fasta ]
```
123

RomainFeron's avatar
RomainFeron committed
124
In this example, `--markers-table` is the table generated with `process` and markers significantly associated with sex are saved to **markers.tsv**. The sex of each individual in the population is given by **popmap.tsv**. Groups of individuals to compare (as defined in popmap) are specified manually with the parameter `--groups`. The minimum depth to consider a marker present in an individual is set to 5, meaning that markers with depth lower than 5 in an individual will not be considered present in this individual.
125

RomainFeron's avatar
RomainFeron committed
126
By default, the `signif` function generates an output file in the same format as the markers depth table. Markers can also be exported to a fasta file using the parameter `--output-fasta`.
127

RomainFeron's avatar
RomainFeron committed
128
The markers table generated by `signif` can be visualized with the `radsex_markers_depth()` function of [sgtr](https://github.com/SexGenomicsToolkit/sgtr), which generates a heatmap showing the depth of each marker in each individual.
129
130


RomainFeron's avatar
RomainFeron committed
131
### Aligning markers to a genome
132

RomainFeron's avatar
RomainFeron committed
133
Markers can be aligned to a genome using the `map` command:
134

RomainFeron's avatar
RomainFeron committed
135
136
137
```bash
radsex map --markers-file markers_table.tsv --output-file alignment_results.tsv --popmap popmap.tsv --genome-file genome.fasta --min-quality 20 --min-frequency 0.1 --min-depth 5 --groups M,F
```
138

RomainFeron's avatar
RomainFeron committed
139
In this example, `--markers-file` is the markers depth table generated with `process` and the path to the reference genome file is given by `--genome-file`; results will are saved to **alignment_results.tsv**. The sex of each individual in the population is given by **popmap.tsv** and the minimum depth to consider a marker present in an individual is set to 5, meaning that markers with depth lower than 5 in an individual will not be considered present in this individual. Groups of individuals to compare (as defined in the popmap) are specified manually with the parameter `--groups`
140

RomainFeron's avatar
RomainFeron committed
141
The parameter `--min-quality` specifies the minimum mapping quality (as defined in [BWA](http://bio-bwa.sourceforge.net/bwa.shtml)) to consider a marker properly aligned and is set to 20 in this example. The parameter `--min-frequency` specifies the minimum frequency of a marker in the population to retain this marker and is set to 0.1 here, meaning that only sequences present in at least 10% of individuals of the population are aligned to the genome.
142

RomainFeron's avatar
RomainFeron committed
143
Alignment results from `map` can be visualized with the `radsex_map_circos()` function of [sgtr](https://github.com/SexGenomicsToolkit/sgtr), which generates a circular plot showing bias and association with sex for each marker aligned to the genome.
144

RomainFeron's avatar
RomainFeron committed
145
Alignment results for a specific contig can be visualized with the `radsex_map_region()` function to show the same metrics for a single contig.
146
147


Romain Feron's avatar
Romain Feron committed
148
## LICENSE
149

RomainFeron's avatar
RomainFeron committed
150
Copyright (C) 2018-2020 Romain Feron
151

Romain Feron's avatar
Romain Feron committed
152
This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation,
153
either version 3 of the License, or (at your option) any later version.
154

Romain Feron's avatar
Romain Feron committed
155
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
156
See the GNU General Public License for more details.
157
158

You should have received a copy of the GNU General Public License along with this program. If not, see https://www.gnu.org/licenses/