“Our study presents the first detailed  performance evaluation of a set of computational methods used to  compare DNA sequences that do not encode proteins -- so-called  "noncoding" sequences -- which make up the majority of sequences in  animal genomes, including our own.”

Our study provides a detailed performance evaluation for a number of widely-used programs in the field of DNA sequence comparison—the alignment of functional noncoding sequences. Identifying conserved noncoding sequences is a commonly used computational approach for detecting regulatory sequences that control gene expression, and many new alignment methods have been developed in the last few years to solve this problem. Until recently, there has been no serious attempt by (or incentive for) developers to evaluate the performance of a new alignment method. We approached the problem of aligning noncoding sequence comparisons from a user's perspective rather than a developer's perspective, giving an assessment that is of interest to a range of researchers, from developers to bench biologists. I believe the real merit of the work is its practical utility—providing an evaluation of a timely methodological question that many biologists face and to which they would like to know the answer.

Computational biology is currently experiencing a rapid growth phase, with large-scale biological datasets and novel computational methods being published at an ever-increasing rate. Choosing the appropriate computational tool for a given biological problem is a situation researchers now regularly encounter, often spending substantial time and effort installing and evaluating different software when they would rather be analyzing their data. Our study presents the first detailed performance evaluation of a set of computational methods used to compare DNA sequences that do not encode proteins, so-called "noncoding" sequences, which make up the majority of sequences in animal genomes, including our own. Noncoding sequences that are conserved between species can predict as-yet-unidentified functionally important sequences such as those that regulate gene expression, and our study helps guide the choice of computational methods to find such sequences.

  How did you become involved in this research?