Detection of non-coding RNA with comparative genomics and the sequential closure of smooth graphs in Cartesian currents

Abstract

In the field of genomics, this thesis presents algorithms for identifying non-coding RNA (ncRNA) genes. It describes a rapid and highly reliable comparative statistical method for identification of functionally significant base pairs in ncRNA genes in multiple sequence alignments of cross-species homologs, a divide-and-conquer approach to optimal assembly of exon predictions with O(n log n) time-complexity, (the standard algorithm for exon assembly has O(n²) time-complexity for ncRNA exon predictions,) and highly accurate statistical tests for exon boundaries based on recognition of non-contiguous patterns in known examples. It also describes a method for scanning cDNA for ncRNA genes. In the field of geometric measure theory, it proves that the set of cartesian currents given by integration over the graphs of smooth functions is dense in the set of all cartesian currents.