Frederick R. Cross | Search bioRxiv

bioRxiv2015

Patching holes in the Chlamydomonas genome

The Chlamydomonas genome has been sequenced, assembled and annotated to produce a rich resource for genetics and molecular biology in this well-studied model organism. However, the current reference genome contains ~1000 blocks of unknown sequence ( N-islands), which are frequently placed in introns of annotated gene models. We developed a strategy, using careful bioinformatics analysis of short-sequence cDNA and genomic DNA reads, to search for previously unknown exons hidden within such blocks, and determine the sequence and exon/intron boundaries of such exons. These methods are based on assembly and alignment completely independent of prior reference assembly or reference annotation. Our evidence indicates that ~one-quarter of the annotated intronic N-islands actually contain hidden exons. For most of these our algorithm recovers full exonic sequence with associated splice junctions and exon-adjacent intron sequence, that can be joined to the reference genome assembly and annotated transcript models. These new exons represent de novo sequence generally present nowhere in the assembled genome, and the added sequence can be shown in many cases to greatly improve evolutionary conservation of the predicted encoded peptides. At the same time, our results confirm the purely intronic status for a substantial majority of N-islands annotated as intronic in the reference annotated genome, increasing confidence in this valuable resource.

Genomics↗

bioRxiv2015

Tying down loose ends in the Chlamydomonas genome

The Chlamydomonas genome has been sequenced, assembled and annotated to produce a rich resource for genetics and molecular biology in this well-studied model organism. The annotated genome is very rich in open reading frames upstream of the annotated coding sequence ( uORFs): almost three quarters of the assigned transcripts have at least one uORF, and frequently more than one. This is problematic with respect to the standard scanning model for eukaryotic translation initiation. These uORFs can be grouped into three classes: class 1, initiating in-frame with the coding sequence (cds) (thus providing a potential in-frame N-terminal extension); class 2, initiating in the 5UT and terminating out-of-frame in the cds; and class 3, initiating and terminating within the 5UT. Multiple bioinformatics criteria (including analysis of Kozak consensus sequence agreement and BLASTP comparisons to the closely related Volvox genome, and statistical comparison to cds and to random-sequence controls) indicate that of ~4000 class 1 uORFs, approximately half are likely in vivo translation initiation sites. The proposed resulting N-terminal extensions in many cases will sharply alter the predicted biochemical properties of the encoded proteins. These results suggest significant modifications in ~2000 of the ~20,000 transcript models with respect to translation initiation and encoded peptides. In contrast, class 2 uORFs may be subject to purifying selection, and the existent ones (surviving selection) are likely inefficiently translated. Class 3 uORFs are remarkably similar to random sequence expectations with respect to size, number and composition and therefore may be largely selectively neutral; their very high abundance (found in more than half of transcripts, frequently with multiple uORFs per transcript) nevertheless suggests the possibility of translational regulation on a wide scale.

Genomics↗