Would it be possible to support `-k Auto`?

[lmrodriguezr]

It is currently possible to select from a set of kits, but I think it should be possible to test all the kits in the first N sequences, and as soon as one of the adapters dominate the others then switch to that. Would that be possible?

I would like to incorporate Ktrim into MiGA, but I want to avoid adding parameters for sequencing technology/kit, and this would be a solution.

[hellosunking]

Dear Luis, Thanks for your interest in my work. I understand that it would be very nice that Ktrim could detect the adapters automatically, and I will try to add this feature in the future versions. However, there are some points that we need to consider: 1. in most scenarios, the authors know the adapter sequences associated with their data, therefore it is not difficult for them to specify them when using Ktrim and this manual step could ensure that the adapters are correctly used; 2. if we want to detect the adapters automatically, the most feasible way is to screen the first N sequences and check a list of known adapters to see which one matches the data. The performance is highly related to the list of known adapters. We need to maintain a list of adapters for popular kits. In addition, if the insert size is long enough then most of the reads contains no, or partial, adapters, the performance of “auto detection” may be affected. Regards, Kun 发件人: Luis M Rodriguez-R <notifications@github.com> 发送时间: 2020年8月24日 7:22 收件人: hellosunking/Ktrim <Ktrim@noreply.github.com> 抄送: Subscribed <subscribed@noreply.github.com> 主题: [hellosunking/Ktrim] Would it be possible to support `-k Auto`? (#6) It is currently possible to select from a set of kits, but I think it should be possible to test all the kits in the first N sequences, and as soon as one of the adapters dominate the others then switch to that. Would that be possible? I would like to incorporate Ktrim into MiGA, but I want to avoid adding parameters for sequencing technology/kit, and this would be a solution. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#6> , or unsubscribe <https://github.com/notifications/unsubscribe-auth/AHEOH5VFL6BO3D2HPO5DJWLSCGQAXANCNFSM4QI5IL6A> .