Issues with resolution

[hw85241]

Hello,

Thank you for generating this useful tool. I'm having trouble running YaHS, particularly involving the following error (which I initially got using the default settings):
"[E::calc_norms] no enough bands (7) for norm calculation, try a higher resolution
[W::run_scaffolding] No enough bands for norm calculation... End of scaffolding round.".

For context, I am working with a contig file that contains 995 contains ranging from 9.2kb to 209.kb in size (~190 are above 50kb). In an attempt to resolve the error, I've both manually designated r ranging across my lowest and greatest contigs:

"yahs -r 9000,10000,50000,100000,200000,500000,1000000,2000000,5000000,10000000,20000000,50000000,100000000,200000000,500000000 OmniC.hap1.fa hap1_unfiltered.bam -o OmniC.hap1_unfiltered_OmniC_hap1”

After increasing the resolution, I then go the memory error: “[I::run_scaffolding] No enough memory. Try lower resolutions... End of scaffolding round.”

I then added the “--no-mem-check” flag using the same resolution settings and got no errors. I should note that memory should not be an issue (I’ve allocated 840Gb). However, the .hic file generated at the end of the suggested commands was very small (531881), and I get an error when trying to load it into Juicebox. I’m assuming this is because there is some issue that’s arisen with YaHS involving the resolution that impacts the downstream steps.

If you could provide any insight on the issues I’m getting, I would greatly appreciate it. Thanks in advance!

[c-zhou]

Hello @hw85241,

If you obtained an AGP file from YaHS and was able to performed downstream analysis to generate a .hic file, the issue is unlikely to be related to resolution. Generating the AGP file with YaHS and creating the .hic file with juicer_tools are separate steps. It may be worth checking for any errors encountered while running juicer_tools during the .hic file construction process.

Regarding the resolution, you have a very small genome and short contigs, so the default resolution setting is not suitable. I recommend trying the following resolution parameter: -r 1000,2000,3000,5000,10000,20000,30000,50000,100000,200000,300000,500000,1000000.

Best,
Chenxi