CRC results reproducibility

#456
by lurcelay - opened

Hi! I am trying to replicate the results of the multi-task learning approach on the colorectal cancer data (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE178341). Using the GSE178341_crc10x_full_c295v4_submit_cluster.csv.gz metadata file I have identified the three classes for MMR status and the 19 classes for cell type classification. Nevertheless, I am unable to identify the 47/53 classes for cell subtypes that you define in the paper.

Could you please let me know the methodology you followed to create the 47/53 cell subtype classes from the metadata file?

Thank you in advance!

Thank you for your question! These were defined by the original authors. Please reach out to the original authors to request the fine labels.

ctheodoris changed discussion status to closed

Sign up or log in to comment