After completing #158 we should run the same two benchmarking samples, SCPCR000372 and SCPCR000373, through the Alevin-fry workflow using the unfiltered mode and add them to our benchmarking results. Here we will want to make sure that we perform filtering on the unfiltered output before performing any comparisons.
When reading in alevin-fry results, only those spots that are also identified with Spaceranger will be included. The next filtering should involve removing any spots that do not include tissue. This can be done by removing any spots where spatialData$in_tissue == 0. Based on the documentation for Spaceranger, it appears that the only filtering they do between the raw and filtered output is by removing spots that don't have tissue overlapping so we should not need any additional filtering beyond that.
Comparisons should be made between Alevin-fry knee, Alevin-fry unfiltered, and Spaceranger by modifying the current benchmarking notebook, analysis/quantifier-comparisons/13-spatial-transcriptomics-benchmarking.Rmd.
After completing #158 we should run the same two benchmarking samples, SCPCR000372 and SCPCR000373, through the Alevin-fry workflow using the unfiltered mode and add them to our benchmarking results. Here we will want to make sure that we perform filtering on the unfiltered output before performing any comparisons.
When reading in alevin-fry results, only those spots that are also identified with Spaceranger will be included. The next filtering should involve removing any spots that do not include tissue. This can be done by removing any spots where
spatialData$in_tissue == 0. Based on the documentation for Spaceranger, it appears that the only filtering they do between the raw and filtered output is by removing spots that don't have tissue overlapping so we should not need any additional filtering beyond that.Comparisons should be made between Alevin-fry knee, Alevin-fry unfiltered, and Spaceranger by modifying the current benchmarking notebook,
analysis/quantifier-comparisons/13-spatial-transcriptomics-benchmarking.Rmd.