Jaltomata QTL mapping

Mapping raw reads from Novogene to tomato genome build 3.0

Checking initial quality of reads with fastqc

for f in $FILES
do
        if [ "$f" != "rawdata_md5.txt" ]
        then
                fastqc $f
        fi
done

Done for all raw read files.

Index tomato genome with BWA

bwa index -a is SLA_3.fa

Trim and clean reads with Trimmomatic

for f1 in *-1_1.fq.gz
do
        f2=${f1%%_1.fq.gz}"_2.fq.gz"
        new=$(echo $f2| cut -d'-' -f 1)
        base="Filtered.fq.gz"
        java -Djava.io.tmp.dir=/scratch/ -Xmx1g -jar /N/soft/rhel6/trimmomatic/0.35/trimmomatic-0.35.jar PE $f1 $f2 -baseout ../cleaned/$new$base ILLUMINACLIP:/N/soft/rhel6/trimmomatic/0.35/adapters/TruSeq3-PE.fa:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:25 MINLEN:36
done

Check quality post-trimmomatic

for f in $FILES
do
        if [ "$f" != "rawdata_md5.txt" ]
        then
                fastqc $f -o 'cleanedQC directory'
        fi
done

Unzip all trimmed files

for f in $FILES
do
        gzip -d $f
done

Map to tomato assembly 3.0

ref='/N/dc2/projects/gibsonTomato/genomes/SLA_3.fa'
for f1 in *_1P.fq
do
        f2=${f1%%_1P.fq}"_2P.fq"
        bwa mem $ref $f1 $f2 -t 32 > $f1-aligned.sam
done

STACKS

STACKS ref_map.pl

We use the BAM files from mapping to tomato genome as input into STACKS (skipping process_radtags)

module load gcc/4.9.2
module load stacks

ref_map.pl -m 3 -S -T 15 -b 1 -A 'BC1' \
        -o /N/dc2/projects/gibsonTomato/jaltomata/rawdata/release_data/stacksOutput \
        -p Jsi6Filtered_1P.fq-aligned.sam.bam \
        -p Jum2Filtered_1P.fq-aligned.sam.bam \
        -r ...<all_individual BAM files>...

STACKS genotypes.pl output to JoinMap format

module load gcc/4.9.2
module load stacks

genotypes -P /N/dc2/projects/gibsonTomato/jaltomata/rawdata/release_data/stacksOutput \
        -b 1 -r 20 -c -m 6 -t BC1 -o joinmap

Convert genotypes to R/QTL format using filter_jp_file.py

Linkage Map

Build linkage map in R using ASMap.

Remove problems markers using correctGenotypes.py

See jaltomataMapping_5.Rmd for details.

Add in trait values

Map QTL in R/QTL

See QTL_scanning.Rmd for details

Candidate gene search

Identifying candidates within QTL regions via blasting marker consensus sequences to tomato.

Identifying loci in peaks for herm nectar color RGB (for all focal traits)

loci <- herm_nec_color_rgb[herm_nec_color_rgb$lod > 2.73,]
loci <- loci[- grep("cL", row.names(a)),]
write.table(loci, sep=",", file="herm_nec_col_RGB.loci")

Gather consensus sequences for these loci

python .\getQTLMarkers.py -v -p .\consensus_seq_jalMap_6_17.fq -q .\herm_nec_col_RGB.loci.txt -o .\

consensus_seq_jalMap_6_17.fq is a fq file with consensus sequences for all loci on the map. getQTLMarkers.py will output fq files for all groups that contain QTL.

Blast the ITAG CDS database

blat /N/dc2/projects/gibsonTomato/genomes/itag_3.2/ITAG3.2_CDS.fasta L.3.consensus.fa output.L3.blast9 -t=
dna -q=dna -out=blast9

Parse blast output

python .\parseBlast.py -v -p .\herm_nectar_color_rgb\output.L3.blast9 -o .\herm_nectar_color_rgb\summaryL.3 -t cds -s .\herm_nectar_color_rgb\61.out.txt