potential issue in get_flipflop_config.py

[slowsmile]

Dear Elizabeth
I noticed your latest script on AAV flipflop identification and thanks for your contribution in the AAV QC field. I tested the script with my data and I found most of my reads are identified as "?" instead of flip or flop. Then I realized that there is a hardcoded score cutoff of 250 in the script and most of the sw score of my reads failed this cutoff. so I wonder if there is a reasonable base for this cutoff or user could actually adjust it?
Also, when query sequence is compared to targets sequence, the user's query sequence plus 10 flanking bases are used for right flip / flops but the entire query sequence is used to compare to left target sequence. I wonder if there is a special reason behind this arrangement?
Thanks a lot
Xiao

[Magdoll]

Hi @slowsmile ,
Sorry for the late response!

First off are you sure you have the same FLIP/FLOP sequences as the one I had in my script? shown here;
https://github.com/Magdoll/AAV/blob/main/get_flipflop_config.py#L16

If your flip/flop sequences are different - or somehow you think yours might be shorter, or have more mutations (?!), yes, I think you can try different scores. I'd start with lowering 250 to 150. But then I'd really recommend visualizing these sequences in IGV to see just how difference your flip/flop regions are.

the user's query sequence plus 10 flanking bases are used for right flip / flops but the entire query sequence is used to compare to left target sequence
I think that might have been a typo - I'll fix it later. thanks!