I have a list of rs ids taken from a GWAS study based on Alzheimers. I have 1000 bp long fasta sequences (both wild and mutant by substituting minor allele at appropriate position). I also have the information related to the same in bed and vcf formats. I'm sorry I meant to say BED file and not BAM file. I've downloaded data specific to my tissue of interest from here in BED format. Like you said, it contains positional events. For each sequence (wild and mutant), I'd like to compute a score based on the overlap of contents of BED file within the sequences. For example, if there exists an SNP in the 4th position of a certain chromosome and I obtain the following: wild sequence : AACTCGG, mutant sequence : AACACGG and the downloaded bed file contains positional events like chr, 1, 4 (fasta for this location : AACT) and chr, 6, 7 (fasta for this location : GG). I want to compute a score for both wild and mutant sequence based on the overlap of bed file on the sequences and use something like [score(wild) - score(mutant)] as a feature for building a classifier. I hope this gives better clarity of my requirement.