Closed amkwong closed 4 years ago
I added small test files for two tissues (Adipose_Subcutaneous and Brain_Cortex), along with their index files, to the data directory. They contain data in the region chr19:448506-528506, and are less than 1 MB each. They will be useful for testing region views on a local machine.
I've done all the updates I can at the moment. There are still some issues I want to tackle (making searchbox.js better, for example), but I'll do it in a new branch. I may still update the buttons in the region view here, but I'm going to leave everything else alone for now. I'm ready for this pull request to be finalized, merged, and updated on the server.
If you're still reading, a quick question to see if you have any good ideas: Is there an easy way to convert the long (ENSG) names to the shorter common gene names in JavaScript that doesn't involve relying on omnisearch or some other outside service? I'm specifically thinking about this: https://github.com/abought/pheget/blob/b3a8594141e5c7f056e0f6f2a69f242d46b79196/pheget/static/js/region.js#L78 JavaScript doesn't seem to like the idea of me loading a file, or else this would be much easier. Is the most straightforward solution to create an API that serves gene symbols when I submit a long gene name?
If you're still reading, a quick question to see if you have any good ideas:
I clearly have work life balance issues this week, so yeah.
You'll generally have issues reading files from JS, and certainly you don't want users to download a big lookup (like a database) on every page view. The network boundary is pretty rough that way.
As for options- I assume this can't be done at page load because it's a dynamic menu. But if the menu already has one name, can we give it two? Check in with me in the AM and we can take a closer look together; I need to look at the data. I'll merge start of the day; I should never merge so near to zonking.
The json file uploaded contains the full conversion table for all genes in GTEx v8. This means we no longer need the python pickle file for gene name lookups, so I removed it, and edited the parts of the code that depends on the pickle.
Ticket
Addresses #33 and #34
Purpose
Deployment / configuration notes
I have already uploaded the necessary files to the deployment server's data directory.
The updated query requires tissue-specific data files when querying from a single tissue. You can find test files at the location below. Note that these files only have data for chr19:448506-528506 for testing purposes.
/net/amd/amkwong/browseQTL/all_chr/data/test.data/Adipose_Subcutaneous.allpairs.sorted.txt.gz /net/amd/amkwong/browseQTL/all_chr/data/test.data/Adipose_Subcutaneous.allpairs.sorted.txt.gz.tbi /net/amd/amkwong/browseQTL/all_chr/data/test.data/Brain_Cortex.allpairs.sorted.txt.gz /net/amd/amkwong/browseQTL/all_chr/data/test.data/Brain_Cortex.allpairs.sorted.txt.gz.tbi
The full data set can be found at the following location: /net/amd/amkwong/browseQTL/all_chr/data/.allpairs.sorted.txt.gz
These files can be generated by the data processing pipeline if you start with raw data from GTEx v8.
Things to fix