Difference Quantitation

The difference quantitation quantitates your data based on the relative levels of reads on different strands.

It can be useful to use the strand of a read to signify some other property of your sequence (eg methylated / unmethylated or exact match vs SNP). This quantitation method then allows you to look for enrichment of this kind of information in a way which accounts for overall biases in the distribution of sequences.

Options

The options you have for this module are:

1. The strands you want to compare (Forward / Reverse / Unknown).
2. How you want to do the comparison (subtract, divide or log divide). If you want to get a simple corrected value you'd probably just subtract one count from another. If you're looking for an enrichment then using a ratio is better. If your enrichment could be either positive or negative you should take a log ratio so that the positive and negative scales are treated the same

For the divide, and log divide options the quantitation will actually add 1 to both of the counts used to avoid getting a divide-by-zero error, so the value you see may be slightly off from what you expect - especially for probes with low absolute counts (from which you probably shouldn't be calcuating this sort of value anyway).