It often occurs in sequencing that one sequencing run does not result in sufficient number of reads for a given analysis, for example due to multiplexing of samples, or due to running a library on a machine with little output like the Illumina MiSeq.
One strategy in these cases is to run a (multiplexed) library multiple times and then combine the results. For example, I did run a multiplexed library of three samples four times on the MiSeq (whole exome sequencing), performed an alignment for each sample for each run and combined the alignments afterwards. However, I wonder to what extend this approach is scalable, especially in RNA sequencing.
Imagine a small RNA-seq experiment (expA) gives you and output of 1 million reads per sample and a small RNA-seq experiment (expB) gives you and output of 10 million reads per sample. Would it be equivalent to perfrom 10 times expA to performing 1 times expB? Regarding the probability to have rare RNAs in your combined output, does this make a difference?
Thank you. Eva
One strategy in these cases is to run a (multiplexed) library multiple times and then combine the results. For example, I did run a multiplexed library of three samples four times on the MiSeq (whole exome sequencing), performed an alignment for each sample for each run and combined the alignments afterwards. However, I wonder to what extend this approach is scalable, especially in RNA sequencing.
Imagine a small RNA-seq experiment (expA) gives you and output of 1 million reads per sample and a small RNA-seq experiment (expB) gives you and output of 10 million reads per sample. Would it be equivalent to perfrom 10 times expA to performing 1 times expB? Regarding the probability to have rare RNAs in your combined output, does this make a difference?
Thank you. Eva