search

wangjr03 / FLAMINGO

MIT License
14 stars 11 forks source link

specific region hic simulation #16

Open poflawless opened 10 months ago

poflawless commented 10 months ago

Hello,

Thanks for your contribution to develop this tool. Could you give me any hints about how to use FLAMINGO to construct 3D organization in specific region? Like 1-2Mb ? And I urgently need this function. Could you help me?

Best Wishes

JiaxinYangJX commented 10 months ago

Thanks for using our tools! Since you just want to reconstruct the structure of a small domain, there is no need to reconstruct it in a hierarchical way as the original paper suggested. Could you provide more information about your Hi-C data, including the format, resolution, size, etc., so that we can know what function can be used. Thanks!

poflawless commented 10 months ago

Thanks for your reply. The format of my hic data is .hic which is generated and normalized by juice box KR method. The resolution is 5KB and libraries were generated with dpnII and DdeI enzyme. Since I am working on plant species, I am not sure which paramaters I should tune. Thanks again for your help

2023-09-26 23:58:25 "JiaxinYangJX" @.***> 写道：

Thanks for using our tools! Since you just want to reconstruct the structure of a small domain, there is no need to reconstruct it in a hierarchical way as the original paper suggested. Could you provide more information about your Hi-C data, including the format, resolution, size, etc.. Thanks!

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

poflawless commented 10 months ago

I have another question which is that in your article, you mentioned that use DNase-seq can predict 3D organization well due to the high correlation between these two data types. Do you have any ideas about how is the relationship between ChIPseq data /ATACseq data/ RNAseq data with hic data? If I want to measure that, which method or tool can I use for the next step?

Best Wishes

JiaxinYangJX commented 10 months ago

For the .hic data, does it contain the entire chromosome or just the specific region like chr1:1MB-2MB? If it contains the entire chromosome, I think the simple idea is just following the reconstruction of the entire chromosome, then you only need to look at the reconstructed 3D coordinates of the specific region.

Remember, when you reconstruct the entire chromosome, the .hic with both high resolution (such as 5kb) and low resolution (such as 1MB) is required.

For the second question, iFLAMINGO used both DNase-seq from the target cell type and Hi-C from the source cell type to reconstruct the 3D structures of the target cell type. Although it shows a high correlation between Hi-C and DNase-seq, we cannot reconstruct the 3D structures using DNase-seq only. If you want to measure the relationship between 1D signals (ChIP-seq / ATAC-seq / RNA-seq) and Hi-C, you can simply do the regression of Hi-C contacts with 1D signals in two fragments and genomic distances.

You can find more information on Supplementary Figure 19.

qdxukang commented 9 months ago

and how about chr1:1MB-2MB?

JiaxinYangJX commented 6 months ago

Hi @poflawless @qdxukang Thanks for using our tools! We just released a lite version of FLAMINGO, which is faster, more memory-efficient, and more user-friendly. We fixed the majority of bugs and hope the new version could help. Below is the link. https://github.com/JiaxinYangJX/FLAMINGOrLite