Abstract
In this review we discuss bioinformatic issues in non-coding RNA analysis, in particular the annotation of genome sequences using covariance models. Some recent innovations for improving the speed and accuracy of covariance models is discussed.