REDItools3 output file

[1127yangyuping]

Hello,Dear author, i have two questions. Question 1: In the output of REDItools3, what does the asterisk (*) in the 4th column (i.e., the Reference field) represent? Question 2: In contrast, in the output of REDItools2, the 4th column is consistently filled with the value 2. What does this value signify, and why is it uniform across all rows?

[ahanden]

It looks like you have misscounted the columns. That is the 4th column, Strand. The asterisk in v3 and the 2 in v2 both represent undetermined strand.

[1127yangyuping]

It looks like you have misscounted the columns. That is the 4th column, Strand. The asterisk in v3 and the 2 in v2 both represent undetermined strand.

Thanks for clarifying that the 4th column, Strand, uses * in v3 and 2 in v2 to indicate undetermined strands. Does the strand direction (+ or -) have any impact on A-to-I mutation analysis? Specifically, if the strand is undetermined, could it affect the accuracy of detecting or interpreting A-to-I mutations?

[ahanden]

Does the strand direction (+ or -) have any impact on A-to-I mutation analysis? Specifically, if the strand is undetermined, could it affect the accuracy of detecting or interpreting A-to-I mutations?

Yes, it absolutely does. If the strand is undetermined, then a reported A-to-G may actually be T-to-C and likewise a reported T-to-C may be A-to-G.

If you are getting a lot of undetermined strand results, you may consider lowering te strand confidence threshold (minimum is 0.5).

But do note, if you're working with a non-stranded RNAseq dataset, you really ought to be using REDItools in unstranded mode. You would then look for both A-to-G and T-to-C in the output.

[1127yangyuping]

Thank you, but I've encountered a problem while processing the results. Could I please help me? When removing SNPs, I've found that using chr:pos:ref:alt to filter SNPs in REDItools results only yields an AG (aggregate match) ratio of around 40%. However, when using chr:pos, the AG ratio reaches over 90%. So, I'd like to ask what filtering method you're using? The SNP files I'm currently using are avsnp147, 1KGP, and gnomAD41. The full match result is still only around 40%.

[ahanden]

How are you providing SNP information? REDItools3 can only filter by base position - entirely excluding a position from the results.