-
Notifications
You must be signed in to change notification settings - Fork 4
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
output file #20
Comments
Using the input file in .cool format should be sufficient. I recommend trying lower resolutions, such as 40kb. If the issue persists, please provide screenshots of any warning messages from DiffDomain and of the output file displaying all columns. |
Thank you for your reply, the last issue has been resolved, which result from our err option. But there is a new problem that has arisen. When I use raw matrix as input I get an error with the following message: During handling of the above exception, another exception occurred: Traceback (most recent call last): The above exception was the direct cause of the following exception: Traceback (most recent call last): I'm guessing this could be a problem with the matrix to cool or hic conversion process, but am not sure exactly why. So I called hicexplorer to convert to cool format and then do the calculations, but it also reported an error. |
python /public/home/xhhuang/miniconda3/envs/diffdomain/lib/python3.7/site-packages/diffdomain_py3/diffdomains.py dvsd multiple J668_mock_0min_Ghjin_D05.cool J668_Fov7_720min_Ghjin_D05.cool Ghjin_D05_TAD_region.bed --reso 20000 |
multiprocessing.pool.RemoteTraceback: The above exception was the direct cause of the following exception: Traceback (most recent call last): |
This is my input file |
I'm guessing if it's due to unequal lines in the input file |
This is the number of hic for the same material at different times |
Since the two matrix matrices did not contain bin interactions equal to 0, the rows were not equal, which made it impossible to compare the two matrices after generating the cool. I used a script in hicpro to generate a symmetric N*N matrix and converted the lower triangular matrix to three columns "bin1 bin2 reads" and then used hicexplorer to convert the matrix to cool and then compared them but the fourth and fifth columns did not have any values.
Convert to cool |
I can get results, but the fifth and sixth columns don't have any values. cool1,cool2, tad and results files in the attachment |
Let's address the issue with this specific usage first. Thank you for providing the example data. Format in For example, the first two lines in the
rather than
Please revise the format of the .matrix files and try this usage again. Kindly let us know if the issue is solved. |
Thank you very much for your answer, I followed your suggestion on bin_id*reso, but still got the error message as below: During handling of the above exception, another exception occurred: Traceback (most recent call last): The above exception was the direct cause of the following exception: Traceback (most recent call last): |
J668_Fov7_720min_Ghjin_D05_reso.zip |
Really think diffdomain is a very useful tool, and thanks for your prompt reply! |
A quick response first. There is a bug when loading three-column files as the input. Current version assigns every chromatin interaction in the input file for every TAD, which is wrong. For example, assigning the interaction Meanwhile, the input file in |
We have fixed the bug in reading chromatin interactions with three-column sparse format. Please follow the instruction in Method1: to install the conda environment to install DiffDomain and rerun the command. This version has been tested on Macos. Full results here |
Thank you for your reply, I would like to ask if the files are normalized (KR or ICE) when using three-column sparse format as input, as I found that a normalization COOL file is generated when using COOL as input, whereas there is no such file when using three-column sparse format as input! |
The three-column sparse format is used as is, with no normalization performed by DiffDomain. |
accroding your suggestion, I want to using the cool file as input file,but generated err: The above exception was the direct cause of the following exception: Traceback (most recent call last): The above exception was the direct cause of the following exception: Traceback (most recent call last): |
the cool file generated by using the hicConvertFormat |
Hi author:
The diffDomain is a useful tool. But there seems to be something strange about my results, all the TADs have no p-values, is it possible that the input file should provide the original matrix not the iced matrix.
thanks.
The text was updated successfully, but these errors were encountered: