I was going to mention spatial transcriptomics. I'...
# microscopyp
Phil Ewels
04/13/2023, 7:57 PMI was going to mention spatial transcriptomics. I'm not really sure what the current state of the project is now, but @Erik Fasterius and @Christophe Avenel have been working on that in the past and I think that there were some plans to extend on into more assay types as it progressed.
f
Florian Wuennemann
04/13/2023, 8:59 PMI am not really sure it would work to extend that pipeline for more assay types, as the data and processing is really different from most imaging based assays. The only other ones that might be able to be implemented in that same pipeline would be sequencing based methods since they probably use sequencing based processing. Most imaging based methods are a lot more focused on image processing.
p
Phil Ewels
04/13/2023, 9:15 PMYeah I was being deliberately vague as I'm trying to think back to discussions that we had several years ago now
Phil Ewels
04/13/2023, 9:16 PMBut I think that the intention was to basically make more of the pure imaging datasets that essentially come for free with these techniques. Do more in the way of image segmentation and annotation etc to get more value that could then be overlayed on top of the spatial sequencing data.
Phil Ewels
04/13/2023, 9:17 PM@Christophe Avenel is the one in this project who came with the most experience in imaging analysis, so I will defer to him though 😬
c
Christophe Avenel
04/14/2023, 1:35 PMHi! For now we decided to focus on Visium data. @Florian Wuennemann is right, we don't plan to extend to more assay types.
About the second part and adding image analysis to down stream analysis, it is still an option for later, but I'm not really sure it's a good fit for that kind of "preprocessing" pipeline.
Regarding file format, we are going for AnnData / H5 format for now, and are waiting for the table specs of NGFF to be adopted a bit before we jump to that.
👍🏻 1