Hello! I am using the methylong pipeline for a few...
# methylong
l
Hello! I am using the methylong pipeline for a few weeks now and I do not understnd how does the it handles barcodes in the pod5 file. I think it is quite common to have all your sequencing data mixed in the same files. But having different barcode corresponding to different experimental condition. Then if one want to make a differerential analysis study (with dmr pair for instance), one need to be able to specify to the program which barcode to select. What I want to do is that each barcode corresponds for instance to a different sample for instance. It is also needed to specify to dorado demux which barcode kit has been used. I do not understand how this is done in your pipeline.)
Copy code
sample,group,path,ref,method,barcode
Project_23_24_DAM_S288c,DAM_S288c,pod5/,ref.fasta,ont,barcode23
Project_23_24_NoDAM_S288c,NoDAM_S288c,pod5/,ref.fasta,ont,barcode24
n
While I am not deeply familiar with the pipeline, I think it does not perform demultiplexing and you would need split your pod5 by barcode and then feed in the individual files as samples. Maybe I am wrong and @Yilia X can comment (@Jin Khoo is currently traveling)
l
Ok, then I am not sure I understand how you perform dmr step.
Sorry I added some details in my previous message.
Thank you for you answer @Niklas Schandry! I do not know any to tool to demultiplex the pod5 files before basecalling them (I am using dorado nanopore data). How would you proceed ? I though the standard use case with dorado was to basecal first and demultiplex afterwards. If this feature is not implemented in the pipeline, would you be interrested in having it implemented ? I maybe I can participate here.
n
I am not maintaining this pipeline, and cannot answer your questions
l
Ok thanks
I checked with collegues I do not think we can demultiplex pod5 because you need at first to have basecalled the data to be able to read the barecodes. To my knowledge, in nanopore sequencing, pod5 contains currents from the pores. To know the sequence from them (including barcodes) you need to basecall.
y
@Léo Tarbouriech Thank you for your question. Currently, methylong doesn't support demultiplexing. It's a tricky issue, and we're discussing and working on it.