Legend issue with Per position quality percentiles plots

[pontushojer]

As you can see from the image below, every line is labeled as a simply black, making it hard to tell the lines apart other than by hovering over them.

Sequail: 1.0.2

Great tool otherwise!

[rhpvorderman]

Thank you for bringing this to my attention. Sorry for the late reply, I am currently recovering from a concussion. So I cannot work on this immediately.

Currently I am of the opinion that the other percentile lines do not add much. They show the spread, sure, but the spread is enormous and follows the same pattern as the mean in almost all cases.

How about I only add the mean and make it green, like the rest of the plots? (Yes, the default colour of all plots is green because I love green. There is no scientific reason for it.)

[pontushojer]

Sorry to hear about your concussion. Hope you have a speedy recovery!

The is no rush from my side, just figured I would raise it for any future developement.

I think dropping the top/bottom 1% would be fine, but to have some idea of the spread would still be useful IMO. Though I guess the "Per position quality score distribution" plots also gives some indication about this.

[rhpvorderman]

Yeah, because they are log scores it is very hard to correctly calculate a good standard deviation. Especially Illumina scores that come only in 4 discrete variants are not properly distributed. ONT is all over the place, it always looks like a pseudo-random number generator to me. Anyway, I will do some research to see if there are proper ways to add error bars here.