Entering edit mode
2.7 years ago
Meghamsh
▴
10
I want to analyse this data in Seurat. However, the format mentioned in the paper is different (.txt file) from what cell ranger usually gives. They have mentioned the version of cell ranger used is 1.2. Since they haven't provided raw fastq, I cannot get the usual 10x format. How to load this data and make a seurat object ?
GEO link for the data: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM3722100
Thank you
When I google "how to lead data into seurat", I find a command to load in text files, so surely you found the same commands and tried them. Can you explain exactly why it won't work for you?
Read10x function by seurat can read data that is in the format of tsv and mtx files. It needs 3 files: Barcodes.tsv, features.tsv and matrix.mtx. Majority of the papers provide their data in either these format or they upload their raw fastq files which can be loaded into cell ranger to get the data in the required format. However, some papers just provide a .txt file where all the rows are genes, all the columns are barcodes and the intersection is the reads. This .txt file, in order to read into a seurat object, needs to be converted somehow. I have tried extracting genes, barcodes into other lists and converting them into tsv and mtx files, but it gives the following error
Error in dimnamesGets(x, value) : invalid dimnames given for “dgTMatrix” object In addition: Warning message: In readLines(con = barcode.loc)
I have looked up the exact search term and I got few forums explaining how to. Its stupid that I assumed that the txt files should be converted in order for seurat to read the data.
https://github.com/satijalab/seurat/issues/4239
Thanks !!