Quantitative measure of clonality for each TCR sequence

[yhf....@gmail.com]

Hi Bo,

I was wondering if there is any metric that could define the number of reads for each unique TCR sequence? Is the 'contig_reads_count' a reliable measure to say that there are more TCR reads for one sequence than the other and thus, there is a bigger number of clonal cells for the one with more contig_reads_counts? I also  read in a previous post back in 2017 where "If you use the est_frequency * DNA sequence length, you will have an MLE estimation of read count". Is this read count a measure of the number of reads for each TCR sequence? Which will be a better metric if I want to determine reads frequency for each TCR sequence?

I do understand that you also mentioned that the 'est_clonal_exp' is an unreliable measure particularly if there is low sequencing depth. Hence, putting it all together, does it suggest this method will be very useful to look at top clonal populations and their CDR3 sequences qualitatively but a limitation will be that we can't fully delineate the quantitative measures of clonality and their frequencies?

Thank you!

Best wishes

Teng

[yhf....@gmail.com]

Hi Bo, 

Sorry to add, but to give my question more clarity, this is essentially what I wanted to explore: I wanted to track the expansion or contraction of  particular T-cell clones over time. and for this, I will be interested to know numbers of each clone for each timepoint experiment from bulk RNAseq data. Hence, I needed some form of TCR frequency measure,and was hoping if you can advise me which metric I can look at? Thanks!

Best wishes

Teng