asl / BandageNG

a Bioinformatics Application for Navigating De novo Assembly Graphs Easily
GNU General Public License v3.0
116 stars 10 forks source link

confusing #120

Closed kingforest93 closed 2 years ago

kingforest93 commented 2 years ago

Dear Dr. Korobeynikov,

Thank you for the development of BandageNG. I am viewing the follwing Hi-C scaffolding results in GFA "Link" and GFA v1.2 "Jump" format in BandageNG. The view of GFA "Link" format is easy to understand, and a linked group of contigs corresponds to a chromosome. But the view of GFA v1.2 "Jump" is confusing, and the linked contigs seem to be randomly placed in the workplace. Could you help explain this problem? Is it due to my incorrect settings or the imperfect design of BandageNG?

GFA format: S chr1_1 LN:i:24433468 S chr1_2 LN:i:5994203 S chr5_1 LN:i:10643942 S chr5_2 LN:i:7618754 S chr5_3 LN:i:8712806 S chr3 LN:i:23459830 S chr2_1 LN:i:8279532 S chr2_2 LN:i:11418757 S chr4_1 LN:i:3274824 S chr4_2 LN:i:5649457 S chr4_3 * LN:i:9660775 L chr1_1 + chr1_2 + 0M L chr5_1 + chr5_2 + 0M L chr5_2 + chr5_3 + 0M L chr2_1 + chr2_2 + 0M L chr4_1 + chr4_2 + 0M L chr4_2 + chr4_3 + 0M

View of GFA format in BandageNG: 5e11d0a9779529d7c75987bece5fce5

GFA v1.2 format: S chr1_1 LN:i:24433468 S chr1_2 LN:i:5994203 S chr5_1 LN:i:10643942 S chr5_2 LN:i:7618754 S chr5_3 LN:i:8712806 S chr3 LN:i:23459830 S chr2_1 LN:i:8279532 S chr2_2 LN:i:11418757 S chr4_1 LN:i:3274824 S chr4_2 LN:i:5649457 S chr4_3 LN:i:9660775 J chr1_1 + chr1_2 + SC:i:1 J chr5_1 + chr5_2 + SC:i:1 J chr5_2 + chr5_3 + SC:i:1 J chr2_1 + chr2_2 + SC:i:1 J chr4_1 + chr4_2 + SC:i:1 J chr4_2 + chr4_3 + * SC:i:1

View of GFA v1.2 format in BandageNG: 54b225e940c3d929826cfa56da48afc

asl commented 2 years ago

Jump links are not included into layout as they could "glue" all the nodes together (they often represent distant links and therefore cannot "guide" the genome).

We might make this behaviour as optional though in the future versions of BandageNG

kingforest93 commented 2 years ago

Thank you for your quick response! I will wait for the future version of BandageNG that can handle Jump links.

Best regards,

Sen Wang

在 2022-10-16 20:54:46,"Anton Korobeynikov" @.***> 写道:

Jump links are not included into layout as they could "glue" all the nodes together (they often represent distant links and therefore cannot "guide" the genome).

We might make this behaviour as optional though in the future versions of BandageNG

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

asl commented 2 years ago

The current version perfectly handles jump links. The decision not to introduce them into layout was made deliberately as these links in general do not provide the proper genomic order and could screw the proper layout otherwise.

kingforest93 commented 2 years ago

Oh, I see. I make a mistake in understanding the design of Jump connections. That is to say, Jump connections should provide correct orders and orientations of contigs just like the Links do. Actually, I'm using GFA Jump to help me view the scaffolding process with Hi-C contact data. This may be already out of the scope of Jump connections in GFA format.

Thank you so much for the patient response! Best regards,

Sen Wang

在 2022-10-16 21:20:42,"Anton Korobeynikov" @.***> 写道:

The current version perfectly handles jump links. The decision not to introduce them into layout was made deliberately as these links in general do not provide the proper genomic order and could screw the proper layout otherwise.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

asl commented 2 years ago

Well, in your particular case you could easily have the following:

J  A  + B +
J  A  + C +
J  B  + C +

Note that the order it correct (genomic), however if you'd add normal links here as well you could easily "glue" everything together. This is why jump links do change the segment layout. It is supposed that you do this by hands.

kingforest93 commented 2 years ago

Thanks for your suggestion! I'll give it a try.

Best regards,

Sen Wang

在 2022-10-16 22:22:35,"Anton Korobeynikov" @.***> 写道:

Well, in your particular case you could easily have the following:

J A + B + J A + C + J B + C +

Note that the order it correct (genomic), however if you'd add normal links here as well you could easily "glue" everything together. This is why jump links do change the segment layout. It is supposed that you do this by hands.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>