Benchmark of long reads mapper using A-lines

[Wenfei-Xian]

          In GFA files,  hifiasm outputs the read IDs and their coordinates if they are used in building assemblies (see A-lines at: https://hifiasm.readthedocs.io/en/latest/interpreting-output.html#interpreting-output).

Originally posted by @chhylp123 in https://github.com/chhylp123/hifiasm/issues/527#issuecomment-1732436001

I am curious can A-lines be used as the ground true in the benchmark of long reads mappers ? Should I used the error corrected reads generated by hifiasm ?

[chhylp123]

Sorry for the late reply. Yes, it might be better to use A-lines. However, hifiasm itself probably has mistakes in both the error correction step and the following assembly step. It is tricky to utilize the corrected reads or not.

[Wenfei-Xian]

thanks !