Lifting over bam

Sometimes for amplicon sequencings, we would like to map reads to the amplicon sequence only but bringing them back to genomic coordinates for easy variant calling and viewing.

Let's say we have a gene in chr1:100-1000, and we would first extract this locus from the genome fasta file to make a new fasta record with name >chr1:100-1000, this can be done with:

echo "chr1:100-1000" | samtools faidx -r - genome.fa > gene.fa 

and map the reads to this single gene fasta file with bwa or bowtie2 to make a bam alignment file:

bwa mem gene.fa query.fq | samtools view -b > gene.bam

So what if you want to put these alignments back to the genomic coordinates after that?

The liftover_bam.liftover function is trying to solve this problem in pure python!

gene_bam="gene.bam"
genome_bam="any.bam_file_that_maps_to_the_genome"
out_bam="where_you_want_your_output_bam_file_to_be"
liftover(gene_bam, genome_bam, out_bam)