Can I please ask the community whether the following is the current best practise for aligning spit paired-end fastq files from Illumina machines obtained from the Sequence Read Archive?
My current pipeline uses the following:
bwa mem -M -R 'readgroupInfo' reference.fa sample1_1.fastq sample1_2.fastq > sample1.sam
The reason I ask is because using this command I receive either errors or odd statistics from downstream processes. For example, when picard is used to mark duplicates, the metrics output contains 0 read pair duplicates and 0 read pair optical duplicates. This seems odd, unless SRA data has already been dedupped?
Also, I am having errors using GATK due to quality scores being too high. My guess is this is because Illumina quality scores have been generated, which can be corrected using the -I flag in the bwa aln algorithm but there doesn't seem to be an equivalent for the mem algorithm. I am currently correcting this in GATK using the --fix_misencoded_quality_scores flag, but could I ask anyone who has experience with this whether this is the correct way to analyse the data.
Many thanks in advance.
My current pipeline uses the following:
bwa mem -M -R 'readgroupInfo' reference.fa sample1_1.fastq sample1_2.fastq > sample1.sam
The reason I ask is because using this command I receive either errors or odd statistics from downstream processes. For example, when picard is used to mark duplicates, the metrics output contains 0 read pair duplicates and 0 read pair optical duplicates. This seems odd, unless SRA data has already been dedupped?
Also, I am having errors using GATK due to quality scores being too high. My guess is this is because Illumina quality scores have been generated, which can be corrected using the -I flag in the bwa aln algorithm but there doesn't seem to be an equivalent for the mem algorithm. I am currently correcting this in GATK using the --fix_misencoded_quality_scores flag, but could I ask anyone who has experience with this whether this is the correct way to analyse the data.
Many thanks in advance.