Question about --use-smith-waterman option

[u1200538]

Hello,
I am glad to post about using the isaac4 aligner.

I am wondering that Isaac aligner --use-smith-waterman smart option.

--use-smith-waterman arg (=smart)
One of the following:

• always : Use smith-waterman to reduce the amount of mismatches in aligned reads
• smart : apply heuristics to avoid executing costly smith-waterman on sequences that are unlikely to produce gaps
• never : Don't use smith-waterman

According to the isaac aligner tool description (https://github.com/Illumina/Isaac4/blob/master/src/markdown/manual.md), when we used the smart option, the mapping position is found by the heuristic method. but if I use the heuristic method, does all the seed length I specified are used for analysis although read doesn't have a gap?

1. I want to know although some reads didn't have a gap, the Isaac aligner uses all of the seed lengths for alignment.
2. If we use the "--use-smith-waterman no" option, does the analysis not use the seed length we specified?

[rpetrovski]

Smart option just uses 7-mers to see if there is a potential to detect gap before running the real Smith waterman. This optimization is not related to seeds. R.

…

On Fri, 13 Aug 2021, 06:47 omG0-hub, ***@***.***> wrote: Hello, glad to posting about using isaac4 aligner. I am wondering that isaac aligner --use-smith-waterman smart option. --use-smith-waterman arg (=smart) One of the following: - always : Use smith-waterman to reduce the amount of mismatches in aligned reads - smart : apply heuristics to avoid executing costly smith-waterman on sequences that are unlikely to produce gaps - never : Don't use smith-waterman In the isaac aligner tool description ( https://github.com/Illumina/Isaac4/blob/master/src/markdown/manual.md), if I used the smart option, the mapping position is found by the heuristic method, but if I use the heuristic method, does all the seed length I specified are used for analysis? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#16>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAXSY4KOFOM3BTGZQ3TUHYDT4SWYLANCNFSM5CC43IDQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .

[u1200538]

Thanks for your help, I understand that the "--use-smith-waterman" cmdline executed before running the Smith waterman.

Additionally, I have some more questions,

1. Any information on what percentage of mismatches or how many base mismatches in the reads does not map?
2. When Isaac's analysis is complete, I'm wondering which criteria classify reads as unmapped reads.
   (For example, I wonder if values ​​like the number of mismatches in the read, base quality or mapping quality are used for classifying unmapped reads.)

[rpetrovski]

Last time I've seen the code was years ago and I'm away from computer for the next few months. From what I remember, there is no mismatch percentage cutoff for mapped reads. The expectation is that high number of mismatches will bring down the mapq. That being said, I think there is a cutoff for accepting rescued mate alignments. R.

…

On Tue, 17 Aug 2021, 12:56 omG0-hub, ***@***.***> wrote: Thanks for your help, I understand that the "--use-smith-waterman" cmdline executed before running the Smith waterman. Additionally, I have some more questions, 1. Any information on what percentage of mismatches or how many base mismatches in the reads does not map? 2. When Isaac's analysis is complete, I'm wondering which criteria classify reads as unmapped reads. (For example, I wonder if values ​​like the number of mismatches in the read, base quality or mapping quality are used for classifying unmapped reads.) — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#16 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAXSY4OOKONPF6BJXJLX243T5I54ZANCNFSM5CC43IDQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .

[u1200538]

Sincerely thank you for your rapid reply,
then, can I think like this?

the high number of mismatches in read >> low level of the Mapping quality >> A read with a mapping quality lower than a certain level (i.e., cutoff) is classified as an unmapped read.

I'm curious as to what values ​​are used to determine unmapped read.