I am working with virus sequences, looking to see if there are unexpected differences in virus genome covverage between samples. in 2 samples, overall expression profiles resemble each other. if there is an area of high coverage from nt 15 to 30, it will be high in both samples, but on a different (y-axis) scale. it may be 500 in sample1, 50 in sample2, with a factor 10. a low cover area at location 50 to 80 nt might be 100 in sample 1 and if the factor stays the same, sample2 would be 10. instead sample2 has a coverage of 20, a lot higher than expected. so I am looking for areas that are over- or under represented. i have what are essentially BED files, but in excel, for each sample. 6 samples represent a susceptible cultivar of plant, 6 a resistant cultivar. The resistant cultivar has far lower coverage, but it does have coverage. trying to see if there areas over or under represented, which might affect virus function. can't use log transform because of zero cover locations. Tools for differential expression analysis are geared towards individual genes and don't work for this. Any thoughts from anyone?
