Articles - Bioinformatics

Comparison of Computational Tools for Alternatively Spliced Transcript Isoforms Quantification

  |   477  |  Post a comment  |  Bioinformatics  |  Genomics, RNA sequencing
This article comments the publication, by Zhang et al. in BMC Genomics 2017, entitled:

Evaluation and comparison of computational tools for RNA-seq isoform quantification.


A total of seven tools were compared: Cufflinks, RSEM, TIGAR2, eXpress, Sailfish[/b, [b]Kallisto and Salmon.

The authors, used RSEM simulated datasets to measure the accuracy of methods, technical replicates of experimental data to test the robustness, and simulated transcripts from the TP53 gene to illustrate the challenges of isoform quantification.

They found that alignment-free methods, such as Salmon, Sailfish and Kallisto, require less computational time while achieving similar or better accuracies compared with other methods.

Cufflinks and eXpress, two alignment-dependent algorithms, have inferior accuracy performance with an RSEM simulated dataset.

TIGAR2 has overall good performance, but the run time and memory requirements render the tool less popular for practical use.

Considering both the accuracy and computational resources needed, Salmonaln and RSEM are the two best performers among the alignment-dependent tools.