Pipeline for mapping RNA-seq reads

It consists of four steps:

1. Downloading reads from ncbi SRA database by SRR id using fastq-dump program
2. Filtering reads using Trimmomatic program
3. Mapping procedure and getting genes counts conducted with STAR
   Warning: gff is not strong file format, so there may be problems with reading different gff files

Requirements:

• Python 2.7
• fastq-dump
• Trimmomatic
• STAR

How to use

Install it:

pip install git+https://github.com/loven-doo/mapipe.git --upgrade  

Have a look at mapipe/configs/config_example.ini to understand how to set fastq-dump, Trimmomatic, STAR and HTSeq in your config.ini file. Look at other example files to know more about mapipe usage

1. If you want prepare only one SRR
   Excute 'mapipe' or 'python -m mapipe' with folowing options:
   -r [--srr]
   -d [--srr_downloads_dir] <Name/of/directory/where/reads/should/be/downloaded>
   -g [--genome_fasta] <Fasta/file/with/chromosomes> (only if indices are not exists)
   -i [--genome_indices] <File/with/genome/indices>
   -a [--gff] <gff/file/with/annotation>
   -c [--config_path] <Your/config.ini/file>

Also you can use import in python:

from mapipe import get_counts

get_counts(srr, srr_downloads_dir, genome_fasta_or_indices, gff, config_path)

2. If you want prepare a list of SRRs
   Execute 'mumapipe' or 'python -m mumapipe' with folowing options:
   -l [--srr_list] <Text/file/with/srr/list>
   -d [--srr_downloads_dir] <Name/of/directory/where/reads/should/be/downloaded>
   -g [--genome_fasta] <Fasta/file/with/chromosomes> (only if indices are not exists)
   -i [--genome_indices] <File/with/genome/indices>
   -a [--gff] <gff/file/with/annotation>
   -c [--config_path] <Your/config.ini/file>
   -t [--threads]

Also you can use import in python:

from mumapipe import run_mapipe

run_mapipe(srr_list, srr_downloads_dir, genome_fasta_or_indices, gff, config_path, threads)

3. You can use mapipe.gff_tools module to manipulate with annotation file (gff)
   3.1. Use 'mapipe.gff_tools.gff3_to_gtf' or 'python -m mapipe.gff_tools.gff3_to_gtf' to convert gff3 to gtf:
   -i [--gff3] <path/to/input/gff3/file>
   -id [--ID] (default: ID)
   -p [--Parent] (default: Parent) -n [--GeneName] (default: gene) -o [--gtf] <path/to/output/gtf/file> (optional)

Also you can use import in python:

from mapipe.gff3_tools import gff3_to_gtf  

gff3_to_gtf(gff, id_attrt='ID', parent_attr='Parent', gene_name_attr='gene', gtf=None)

4. If you want conduct only one step use import in python:

import mapipe  

mapipe.download_reads(srr, srr_downloads_dir, config_path) # downloads reads from SRA database using fastq-dump

reads_dir = os.path.join(srr_downloads_dir, srr) # joins the path to fastq files
mapipe.filter_reads(reads_dir, config_path) # filters reads using Trimmomatic

mapipe.map_reads(reads_dir, genome_fasta_or_indices, gff, config_path) # runs STAR (it can automatically index input genome)  
# If you included gff file into genome indices you can use gff=None  
# If you do not want to calculate gene counts for input reads add calc_counts=False to the map_reads function call

mapipe.index_genome(genome_fasta, gff, config_path, genome_indices)
# Indexes input fasta file with chromosomes, genome_indices="Genome_indices" by default
# so you can change destination folder name  
# If you do not want include gff file into genome indices use gff=None

If you have any questions or suggestions use this adress to contact: loven-doo@fbb.msu.ru