search

yangao07 / abPOA

abPOA: an SIMD-based C library for fast partial order alignment using adaptive band
MIT License
118 stars 18 forks source link

Feature request: Allow specification of full scoring matrix #20

Closed glennhickey closed 3 years ago

glennhickey commented 3 years ago

Only a match score and mismatch penalty are available now. Would it be possible to allow specification of a full matrix? We would like to be able to, for example, use the default lastz scoring matrix referred to here: http://www.bx.psu.edu/~rsharris/lastz/README.lastz-1.04.03.html#options_scoring

As far as I can tell, this is just requires additional interface given that a full matrix seems to be used internally? Personally, I'm more interested in doing this via the C API than the command line, but I'm sure other users would like the option there too. Thanks!

glennhickey commented 3 years ago

On a slightly related note: I've noticed that abpoa can give quite different results when I do things like scale all the parameters up by the same value. Is this by design? Are there any maximums or hidden paramters that can affect this?

yangao07 commented 3 years ago

scale all the parameters up by the same value

How did you scale the parameters? Did you only scale the match/mismatch or all the scoring parameters?

Yan

yangao07 commented 3 years ago

Only a match score and mismatch penalty are available now. Would it be possible to allow specification of a full matrix? We would like to be able to, for example, use the default lastz scoring matrix referred to here: http://www.bx.psu.edu/~rsharris/lastz/README.lastz-1.04.03.html#options_scoring

As far as I can tell, this is just requires additional interface given that a full matrix seems to be used internally? Personally, I'm more interested in doing this via the C API than the command line, but I'm sure other users would like the option there too. Thanks!

This should be easy to update. I will work on that.

glennhickey commented 3 years ago

Thanks!

About the scaling being inconsistent: I was unable to reproduce an example to send you. I must have been mistaken when I thought I saw this happen. If I can find an example, I'll send it right away. In the meantime, sorry!

yangao07 commented 3 years ago

Check out the latest commit of abpoa (v1.2.2), which now takes a new parameter -t score.matrix.

glennhickey commented 3 years ago

Thanks!!!