Longer Reads Do Not Significantly Improve RNA-seq Results.

The initial next-generation sequencing technologies produced reads of 25 or 36 base-pairs and only read from a single-end of the library sequence. Currently, it is possible to reliably produce 300bp paired-end sequences for RNA-expression analysis. While these read lengths have consistently increased, people have assumed that longer reads are better and that paired-end reads produce better results than single-end reads. These assumptions have been based upon intuition rather than hard experimentation. Using the RNA-seq standards from the Association of Biomolecular Facilities – Next Generation Sequencing (ABRF-NGS) Study, we were able to evaluate the impact of read-length on RNA-seq results. We started with paired-end 100bp reads and then trimmed them to simulate different read lengths along with separating the pairs to produce single-end reads. For each read-length and paired status, we evaluated differential expression levels between two standard samples and compared the results to those obtained by qPCR. We found that with the exception of reads trimmed to 25bp, there is little difference for the detection of differential expression regardless of the read-length. Once single-end 50bp reads are used, the results do not change substantially for any level up to and including 100bp paired-end reads. Thus, a researcher could save substantial resources by using 50bp single-end reads for their paired-end expression analysis. We replicated these results by using multiple computational pipelines to confirm that they were not a result of the particular algorithm we were using. Additionally, we performed the same analysis on two ENCODE samples and found consistent results affirming that our conclusions have broad application.