# This file is part of LRTrans which is a toolset for clustering
# and filtering the aligned long-read transcript sequences
#
# LRTrans include three python scripts:
# (1)bam_parser (2)read_cluster (3)transcript_filter
#
#
# These scripts are 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.
#
# Contact: Dr. Shi-Yi Chen, sychensau@gmail.com
# STEPS
After preprocessing steps such as quality control and error correction,
long cDNA/RNA sequences were aligned against the reference genome
using Minimap2 or the like tools. With samtools, the alignments can be
saved in a sorted BAM file. Starting with the sorted bam file,
LRTrans can construct and quantify transcript isoforms by performing
the following steps.
1. Parse and extract information from SORTED BAM file.
Tool:
bam_parser
Command:
./bam_parser -i sorted.bam -r reference.fa -o parsed_read.tsv
Usage:
bam_parser --sorted_bam_file path_to_bam --reference_file path_to_fasta --parsed_file path_to_output
Input files:
-i , --sorted_bam_file: Path to the input bam file. (Required)
-r , --reference_file: Path to the reference fasta file used to extact splice site sequences. (Required)
Output file:
-o , --parsed_file: Path to the output file required by the cluster program. (Required)
2. Cluster transcript reads into isoforms. The abundance of each isoform will be recorded.
Tool:
read_cluster
Command:
./read_cluster -i parsed_read.tsv -o clustered_read.tsv
Usage:
read_cluster --parsed_file path_to_tsv --cluster_file path_to_output
Input files:
-i , --parsed_file: Path to the tsv file outputted by bam_parser. (Required)
Output file:
-o , --cluster_file: Path to the output file that include cluster information. (Required)
Cluster parameter:
-ebt, --exon_boundary_tolerance Three prime exons boundary tolerance. (default = 15)
-ibt, --introns_boundary_tolerance Internal introns boundary tolerance. (default = 3)
-msi, --minimum_size_of_intron Introns with a length (bp) below this value are ignored. (default = 30)
3. Quality filtering and output transcriptoms(GFF3) and representative sequence(fasta).
Tool:
transcript_filter
Command:
./transcript_filter -i clustered_read.tsv -b sorted.bam -op filter_out
Usage:
transcript_filter --cluster_file path_to_tsv --bam_file path_to_bam --out_prefix path_to_out
Input files:
-i , --cluster_file: Path to the tsv file outputted by read_cluster. (Required)
-b , --bam_file: Path to the input bam file. (Required)
Output file:
-op, --out_prefix Prefix/path to the output files. (Required)
filter parameter:
-dn , --standard_donor Standard donor of splice sites, set off with commas.(default = GT,GC,AT)
-ac , --standard_acceptor Standard acceptor of splice sites. (default = AG,AC)
-ia , --isoform_abundance Isoform whose abolute abundance is below this value will be filtered. (default = 3)
-sa , --single_exon_abundance Isoform with single exon will be filtered if its absolute abundance is lower than this value. (default = 3)
-fr , --filter_by_ratio The minimum ratio of the isoform relative abundance in the parent gene locus in clusting. (default = 0.05)
-ej , --erroneous_junction Isoform whose number of erroneous junction sites is above this value will be filtered. (default = 0)