Error occured when Generating Fasta after JBAT Review

[shen Eric]

Hi!

   I manually assembled and generated final fasta and hic file
   but, got error below when generate .hic file. I didn't get hic file.

   And the heatmap generated by remapping to new draft genome did not match the heatmap I manually adjusted.

:( Assembly file does not match cprops file.

[Olga Dudchenko]

Hi Eric,

Are you running the run-pipeline-post-review.sh script? See page 39-40 of the manual http://aidenlab.org/assembly/manual_180322.pdf for more details.

Best,

Olga

[shen Eric]

  Hi Olga,

 

  Yes, I'm running this script.

 

  And I can use the review.assembly file by "import modified assembly" to visualize the final heatmap after import the original assembly.

  thanks,

  Eric

[Olga Dudchenko]

Hi Eric,

One thing I can think of is carriage return characters: we've seen this happen on Windows machines. We have safeguarded against this but have not yet released the fix. Try to check if you have any unusual end-of-line characters (e.g. in Notepad: https://superuser.com/questions/409919/cr-and-lf-characters-shown-at-the-end-of-each-line-in-notepad) or just try

tr '\r' '\n' < review.assembly > new.review.assembly

and see if this fixes the problem.

Best,

Olga

[shen Eric]

  Hi Olga,

  Thanks for your help!

  Now I get the .hic file, but its heatmap is not what I manually adjusted before. It's still erroneous but not the same as original heatmap.

  Best,

  Eric

[Olga Dudchenko]

Hi Eric,

Just to be sure: loading modified assembly is a bit tricky in JBAT. Are you certain that the map you are comparing to is to, in JBAT, load map -> original map, load modified assembly map -> load your review file.. I am referring to this section of the manual:

"

Assembly menu can also be used to load saved changes (menu item “Import Modified Assembly”). Note that when loading changes after exiting the application it is paramount to first load the original (unmodified) .assembly file via the “Import Map Assembly” menu item, and only then load the modified .assembly file. This is because, at least at present, the .hic file by itself does not contain any information about how the contact matrix positions relate to underlying sequences. The information needs to be added before the changes encoded in ,review.assembly file can be meaningfully interpreted.

"

We know this execution is not most intuitive and we've have run into problems here where people would accidentally mix the original and the review up, e.g. load the review as the original file and proceed. So wanted to double check before trying to look for other potential problems. In principle if you load your original and then load your modifications as a modified assembly, it should match what you will get by running finalize and thus rebuilding the .hic file from scratch.

Hope this helps,

Olga

[shen Eric]

  Hi Olga,

  I find the problem. I used the wrong orig assembly file to mannually adjust the heatmap and of course the review assembly is wrong.

  I used the resolved.hic file and the final.assembly file, which is not paired.

 

  I think I should use rawchrom.hic and polished.split.assembly file.

  Thanks for your help!

  Best,

  Eric

[Olga Dudchenko]

Hi Eric,

Glad to hear! Just as a note, we typically recommend using .final.hic and .final.assembly for review. If you are running in haploid mode .final is the same as .rawchrom. Hope this helps,

Olga