LongReadAnnotation

Project Description

LongReadAnnotation is a tool for annotating and validating long reads from sequencing data. It includes pipelines for identifying start and end site clusters and validating the long reads.

Citation

Erik K Flemington, Dinh Truong Nguyen

Requirements

• Perl 5

Dependencies

• File::Basename

Installation

1. Clone the repository:

git clone https://github.com/flemingtonlab/LongReadAnnotation.git

2. Install the dependencies (ensure Perl and required modules are installed).

File Formats

File	Description
Wiggle	Standard wiggle format
BED12	Standard BED12 format
BED6	Standard BED6 format
Genome fasta	Standard genome fasta (can be wrapped or unwrapped)

LongReadAnnotation Pipeline

1. Identify Start Site Clusters

1. Call 5' peaks from wig files (use “-CAGE y” option which shifts output 1bp downstream for CAGE data because STAR aligner CAGE output option (—outWigType read1_5p) outputs signals 1bp upstream from start of read) :

   perl /PATH/peak_caller_from_wigs.pl -w PATH/MC1_Unique.str1.out_chr1.wig,PATH/MC2_Unique.str1.out_chr1.wig,PATH/MC4_Unique.str1.out_chr1.wig -mw 8 -fva 0.2 -mspd 10 -s + -CAGE y

2. Negative strand input wigs values must be negative! (If coverage values for negative strand wigs are positive, use invert_wigs.pl to change sign):

   perl /PATH/peak_caller_from_wigs.pl -w PATH/MC1_Unique.str2.out.negative_values_chr1.wig,PATH/MC2_Unique.str2.out.negative_values_chr1.wig,PATH/MC4_Unique.str2.out.negative_values_chr1.wig -mw 8 -fva 0.2 -mspd 10 -s - -CAGE y

3. Concatenate positive and negative strand bed files:

   cat /PATH/CAGE_peaks_positive_strand.bed /PATH/CAGE_peaks_negative_strand.bed > /PATH/CAGE_peaks_positive_plus_negative_strand.bed

4. Sort concatenated positive plus negative strand bed files (not essential but is a good configuration for the file)

   sort -V -k 1,1 -k 2,2n -k 3,3n /PATH/CAGE_peaks_positive_plus_negative_strand.bed > /PATH/CAGE_peaks_positive_plus_negative_strand_sorted.bed

2. Identify 3' Site Clusters

1. 3’ LR end summary (output is wiggle file coverage of 3’ end coverage):

   perl /PATH/LR_3prime_end_summation.pl -bed /PATH/LR.bed

2. Identify 3' end clusters:

   perl /PATH/peak_caller_from_wigs.pl -w PATH/MC1_3p_chr1.wig,PATH/MC2_3p_chr1.wig,PATH/MC4_3p_chr1.wig -mw 8 -fva 0.2 -mspd 10 -s +
   perl /PATH/peak_caller_from_wigs.pl -w PATH/MC1_3p_negative_values_chr1.wig,PATH/MC2_3p_negative_values_chr1.wig,PATH/MC4_3p_negative_values_chr1.wig -mw 8 -fva 0.2 -mspd 10 -s -

3. Concatenate positive and negative strand wig files

   cat /PATH/3p_peaks_positive_strand.bed /PATH/3p_peaks_negative_strand.bed > /PATH/3p_peaks_positive_plus_negative_strand.bed

4. Sort concatenated positive plus negative strand wig files (not essential but is a good configuration for the file)

sort -V -k 1,1 -k 2,2n -k 3,3n /PATH/3p_peaks_positive_plus_negative_strand.bed > /PATH/3p_peaks_positive_plus_negative_strand_sorted.bed

3. Validate Long Reads

1. Long read validation:

   perl /PATH/LR_validate.pl -5Pp /PATH/MU_5P_CAGE_peaks_chr1.bed -mcde 10 -mcdi 2 -3Pp /PATH/MU_3P_peaks_chr1.bed -3Pde 10 -3Pdi 10 -minSJ 1 -SJt /PATH/MU-SJ.out.tab -f /PATH/hg38_chr1_first_portion.fa -LR /PATH/MU_LR_fullLength.merged_1million.bed

2. Long read validation with previously identified ATG start sites enforcing ORF predictions:

   perl /PATH/LR_validate.pl -5Pp /PATH/MU_5P_CAGE_peaks_chr1.bed -mcde 10 -mcdi 2 -3Pp /PATH/MU_3P_peaks_chr1.bed -3Pde 10 -3Pdi 10 -minSJ 1 -SJt /PATH/MU-SJ.out.tab -f /PATH/hg38_chr1_first_portion.fa -LR /PATH/MU_LR_fullLength.merged_1million.bed -ATG /PATH/LR_validate/test_data/hg38_chr1_known_ORF_start_sites.bed