Reads not being used with lower uniting abundance

[skoren]

I ran MBG v1.0.14 with the same input corrected HiFi reads, the first run used --unitig-abundance 2 while the second one used --unitig-abundance 4. The run with the higher abundance has an example of about 30 reads which are completely missing in the other run. These reads are near the end of the chromosome and telomeric and extend the assembly by about 10kb. The reads are on globus under MBG_ missing_reads folder.