JaneliaSciComp / multifish

EASI-FISH analysis pipeline for spatial transcriptomics
BSD 3-Clause "New" or "Revised" License
32 stars 13 forks source link

Support for dax file input format #7

Open pranathivemuri opened 2 years ago

pranathivemuri commented 2 years ago

New feature

Hi @krokicki! Thanks for the very well-written repo. wondering if you would add support for dax file format images that are obtained from imaging rounds from MERFISH? It seems like the pipeline currently takes in .czi and .mvl files as input

Proposal

  1. Dax file support for stitching module

Usage scenario

It could be used for the analysis of MERFISH data

Suggest implementation

krokicki commented 2 years ago

Hi @pranathivemuri, thanks for the suggestion, we'll definitely take it into account for a future version.

pranathivemuri commented 2 years ago

Hi @krokicki thanks for your reply. If I were to implement it, would it require changing the first module i.e registration module to be compatible, or do other steps of the pipeline expect czi, mvl formats in any way? thanks in advance for your reply!

krokicki commented 2 years ago

@pranathivemuri, it would only require a change to the stitching module, specifically the code here: https://github.com/saalfeldlab/stitching-spark and here: https://github.com/JaneliaSciComp/multifish/blob/master/workflows/stitching.nf Once the data is ingested and stitched, the rest of the pipeline works from that stitched N5. You might want to create an issue over in that stitching-spark repo and ask those developers how difficult it would be to add dax support.

By the way, as a workaround, if you have a different stitching process you can use for your data, you could just find a way to convert your final data into N5 format, and place that into the expected locations (e.g. outputs/acq_name/stitching/export.n5), and then use --skip stitching to run the pipeline from that point forward.

There are many ways to convert data to N5 format. If you can get your data into a TIFF series, there are two Nextflow-based TIFF to n5 converters here: https://github.com/JaneliaSciComp/nf-demos/blob/main/n5-tools-dask/tiff-to-n5.nf https://github.com/JaneliaSciComp/nf-demos/blob/main/n5-tools-spark/tiff-to-n5.nf