Num reads to scan for read length/insert length/mapq

[danoreper]

Hi Lisle,
I'm applying Abra2 to rna data in which the read lengths vary dramatically from the first record to the last.
I noticed in the Abra logs that it had underestimated read length; it seems that this was caused by a hardcoded number of reads to scan in Realigner.java (See line 1455)

Would it be possible to allow the user to pass in the number of reads to scan for estimating the various read-specific quantities? Or perhaps a boolean indicating that all reads should be scanned?
Thanks,
Dan