Skip to content

Latest commit

 

History

History
246 lines (185 loc) · 8.07 KB

13b_GenePrediction_PreparatoryAssembliesAndPredictions.md

File metadata and controls

246 lines (185 loc) · 8.07 KB
Raw

Prepare all input assemblies and predictions for gene prediction

Trinity


#Trinity will not run in a satisfactory time with the rDNA reads, so running with them removed.
ln -s ../11_AlignRNA/SCNgenome.fasta
ln -s ../20_ribosomalArrayID/NonRiboReads.bam

sh runTrinity.sh NonRiboReads.bam


#runTrinity.sh
################################################################################
#!/bin/bash

module load GIF2/trinity

bam="$1"
out=$(basename ${bam%.*} |cut -f 1 -d "_")
Trinity \
   --genome_guided_bam ${bam} \
   --max_memory 110G \
   --genome_guided_max_intron 30000 \
   --full_cleanup \
--CPU 16
################################################################################

#finished in 32hrs
#how many transcripts did we get?
grep -c ">" Trinity-GG.fasta
110683

#map trinity transcripts back to genome
/work/GIF/remkv6/Baum/04_Dovetail2Restart/12_Trinity/trinity_out_dir

sh runGmap.sh SCNgenome /work/GIF/remkv6/Baum/04_Dovetail2Restart/12_Trinity/trinity_out_dir/ ../SCNgenome.fasta Trinity-GG.fasta

How many gene entries did we get in the gff?
less SCNgenome.Trinity-GG.gff3 |awk '$3=="gene"' |grep "path1"|wc
110013  990117 12716727


[remkv6@condo075 23_Mikado]$ less SCNgenome.Trinity-GG.gff3|grep "path1" | awk '$3=="mRNA"{if($4>$5) {print $4-$5} else {print $5-$4}}'  |summary.sh
Total:  702,181,499
Count:  110,013
Mean:   6,382
Median: 563
Min:    8
Max:    1,725,421
[remkv6@condo075 23_Mikado]$ less SCNgenome.Trinity-GG.gff3 |grep "\.mrna1\." |awk '$3=="exon"{if($4>$5) {print $4-$5} else {print $5-$4}}'  |summary.sh
Total:  94,005,020
Count:  499,230
Mean:   188
Median: 144
Min:    1
Max:    6,096

set up braker on unmasked genome

#/work/GIF/remkv6/Baum/04_Dovetail2Restart/13_braker

#note this was ran with all rRNA included
ln -s ../11_AlignRNA/AllRNASEQ_sorted.bam
ln -s ../11_AlignRNA/SCNgenome.fasta
ln -s ../11_AlignRNA/SCNgenome.consensus_isoforms_sorted.bam
ln -s ../11_AlignRNA/SCNgenome.H.glycinesEST_sorted.bam

module use /work/GIF/software/modules
module load GIF/braker/2.1.0
braker.pl --species=Hglycines2 --genome=SCNgenome.fasta --bam=SCNgenome.consensus_isoforms_sorted.bam,SCNgenome.H.glycinesEST_sorted.bam,AllRNASEQ_sorted.bam

#how many genes?
grep -v "#" augustus.hints.gff |awk '$3=="gene"' |wc
  35514  319626 2025377

#how many transcripts?
awk '$3=="transcript"' augustus.hints.gff |grep -v "#" |wc
 37897  341073 2502318

run class2 to generate transcripts

/work/GIF/remkv6/Baum/04_Dovetail2Restart/15_class2
ln -s ../11_AlignRNA/AllRNASEQ_sorted.bam
ln -s ../10_RepeatModeler/SCNgenome.fasta

module load GIF/class2
run_class.pl -a AllRNASEQ_sorted.bam -o AllRNASEQClass2.gtf -p 16 --verbose --clean

#how many transcripts did we generate?
less transcripts.gff |awk '$3=="transcript"' |wc
  60967  853538 8375727

  less AllRNASEQClass2.gtf  |awk '$3=="transcript"{if($4>$5) {print $4-$5} else {print $5-$4}}'  |summary.sh
  Total:  448,279,477
  Count:  60,967
  Mean:   7,352
  Median: 1,985
  Min:    23
  Max:    392,100
  [remkv6@condo075 23_Mikado]$ less AllRNASEQClass2.gtf  |awk '$3=="exon"{if($4>$5) {print $4-$5} else {print $5-$4}}'  |summary.sh      Total:  85,792,277
  Count:  441,887
  Mean:   194
  Median: 149
  Min:    1
  Max:    13,732

set up stringtie

#/work/GIF/remkv6/Baum/04_Dovetail2Restart/16_Stringtie
#Tried to run class2 with all of the rnaseq, but times out without finishing, so used rrna removed bam.


ln -s ../10_RepeatModeler/SCNgenome.fasta
ln -s ../20_ribosomalArrayID/NonRiboReads.bam

module load stringtie
stringtie NonRiboReads.bam -j 5 -p 16 -v -o NonRrnaRNASEQ_stringtie.gtf

#how many transcripts did we generate?
less NonRrnaRNASEQ_stringtie.gtf |awk '$3=="transcript"' |wc
 41504  747072 6605478

 #get transcript and exon stats since there are no genes predicted here

awk '$3=="transcript"{if($4>$5) {print $4-$5} else {print $5-$4}}' NonRrnaRNASEQ_stringtie.gtf |summary.sh
Total:  233,044,848
Count:  41,504
Mean:   5,614
Median: 2,173
Min:    199
Max:    523,622

awk '$3=="exon"{if($4>$5) {print $4-$5} else {print $5-$4}}' NonRrnaRNASEQ_stringtie.gtf |summary.sh      Total:  70,045,713
Count:  277,955
Mean:   252
Median: 153
Min:    3
Max:    15,070

Set up portcullis to analyze splice junctions

#/work/GIF/remkv6/Baum/04_Dovetail2Restart/17_Portcullis
#This initially on the bam file with the rdna reads. Even after removing the rRNA reads, I had to run this on a high mem node, used ~180GB ram.

ln -s ../10_RepeatModeler/SCNgenome.fasta
ln -s ../20_ribosomalArrayID/NonRiboReads.bam

module load portcullis
portcullis full --threads 9 --verbose --use_csi --output portcullis_out --orientation FR SCNgenome.fasta NonRiboReads.bam

Set up braker on a masked genome, excluding the simple repeats

#/work/GIF/remkv6/Baum/04_Dovetail2Restart/19_brakerMasked
ln -s ../11_AlignRNA/AllRNASEQ_sorted.bam
ln -s ../13_braker/SCNgenome.consensus_isoforms_sorted.bam
ln -s ../13_braker/SCNgenome.H.glycinesEST_sorted.bam
ln -s ../10_RepeatModeler/SCNgenome.fasta.masked

module load GIF/braker/2.1.0
braker.pl --species=Hglycines3 --genome=SCNgenome.fasta.masked --bam=SCNgenome.consensus_isoforms_sorted.bam,SCNgenome.H.glycinesEST_sorted.bam,AllRNASEQ_sorted.bam

#how many genes?
awk '$3=="gene"' augustus.hints.gff |grep -v "#" |wc
  22408  201672 1272920

#how many transcripts
awk '$3=="transcript"' augustus.hints.gff |grep -v "#" |wc
  24481  220329 1610884

#translate the gff to fasta for use with spades below
~/common_scripts/gff2fasta.pl ../../../10_RepeatModeler/SCNgenome.fasta augustus.hints.gff 4SPADES

Since cufflinks is deprecated to stringtie, I decided to use another assembler for transcripts

#/work/GIF/remkv6/Baum/04_Dovetail2Restart/22_spadesTranscripts
ln -s ../10_RepeatModeler/SCNgenome.fasta
ln -s ../20_ribosomalArrayID/NonRiboReads.bam
ln -s ../19_brakerMasked/braker/Hglycines3/4SPADES.cdna.fasta

#extract only those reads that aligned to the genome
module load picard/2.17.0-ft5qztz; java -Xmx120G -Xms50G -jar  /opt/rit/spack-app/linux-rhel7-x86_64/gcc-4.8.5/picard-2.17.0-ft5qztzntoymuxiqt3b6yi6uqcmgzmds/bin/picard.jar SamToFastq I=NonRiboReads.bam FASTQ=NonRiboReads_R1.fq F2=NonRiboReads_R2.fq FU=NonRiboReads_unpaired.fq

#Picard failed on some of the reads, due to a mate being removed in ribosomal areas.  ignored this error.

#run spades with extracted reads and using trusted contigs from the masked braker gene prediction
module load spades; spades.py --trusted-contigs 4SPADES.cdna.fasta --rna -m 120 -t 16 -1 NonRiboReads_R1.fq -2 NonRiboReads_R2.fq -o SpadesAssembly

Spades Assembly

#/work/GIF/remkv6/Baum/04_Dovetail2Restart/22_spadesTranscripts

SCNgenome.fasta -> ../10_RepeatModeler/SCNgenome.fasta
NonRiboReads.bam -> ../20_ribosomalArrayID/NonRiboReads.bam
4SPADES.cdna.fasta -> ../19_brakerMasked/braker/Hglycines3/4SPADES.cdna.fasta

#used the cdna's predicted from braker using isoseq, est's, and rnaseq with a masked genome.
SPAdes-3.13.1-Linux/bin/spades.py --trusted-contigs 4SPADES.cdna.fasta --rna -m 300 -t 32 -1 NonRiboReads_R1.fq -2 NonRiboReads_R2.fq -o SpadesAssemblyFat

#/work/GIF/remkv6/Baum/04_Dovetail2Restart/22_spadesTranscripts/SpadesAssemblyFat
grep -c ">" transcripts.fasta
100402

#need to align these so they are available as a gff for mikado
/work/GIF/remkv6/Baum/04_Dovetail2Restart/22_spadesTranscripts/SpadesAssemblyFat/01_GmapAlign

sh runGmap.sh SCNgenome /work/GIF/remkv6/Baum/04_Dovetail2Restart/22_spadesTranscripts/SpadesAssemblyFat/01_GmapAlign/ SCNgenome.fasta transcripts.fasta

less SCNgenome.transcripts.gff3 |awk '$3=="gene"' |grep "path1"|wc
  70737  636633 10925032


#get transcript and exon stats since no genes are predicted -- aligned with gmap so only counts for primary alignments here
[remkv6@condo075 23_Mikado]$ less SCNgenome.transcripts.gff3 |grep "path1" |awk '$3=="mRNA"{if($4>$5) {print $4-$5} else {print $5-$4}}
'  |summary.sh
Total:  858,591,281
Count:  70,737
Mean:   12,137
Median: 795
Min:    2
Max:    1,781,465

less SCNgenome.transcripts.gff3 |grep "\.mrna1\." |awk '$3=="exon"{if($4>$5) {print $4-$5} else {print $5-$4}}'  |summary.sh
Total:  63,693,622
Count:  389,188
Mean:   163
Median: 114
Min:    1
Max:    9,438