smallgenes

Simple eukaryotic genes in small genomic contexts.

Background

Genes and genomes can vary in size from small to HUGE. Smaller genes are useful for development, testing, and tutorials. This project seeks to create a set of small genes primarily used in the "isoforms project".

A smallgene locus has the following features:

• The DNA of the locus is represented in a single FASTA file
• The annotation is represented in a single GFF3 file
  • gene, mRNA, exon, UTRs, and spliced RNA-Seq are kept
  • all other features are removed
• There is only 1 gene in the region (no gene overlaps, not even ncRNAs)
• The gene is flanked by a little genomic sequence (e.g. 99 bp)
• The gene is protein-coding
• At least 1 transcript is spliced
• The gene has canonical features
  • Introns follow the GT-AG rule
  • CDS has normal start and stop codons
  • The intron and exon lengths are not unusually short
• The gene is not too long (e.g. <= 1000 bp)
• The gene is moderately to highly expressed (e.g. > 10,000 observed splices)
• The splicing landscape isn't too complex (e.g. < 10M possible isoforms)
• Some features are purged from the GFF
  • All features in the flanking region are removed
  • RNA-Seq introns with non-canonical splice sites are removed
  • RNA-Seq introns from the opposite strand are removed
• The gene is not too similar to other genes in the set
• The name of the locus follows g.i.r where
  • g is an abbreviation for the genome
  • i is the number of introns
  • r is the rank of splicing complexity compared to others in i

Genomes

The initial build is the C. elegans genome. Other genomes will eventually follow. The build procedure for each genome may be a little different. Details for the C. elegans build are in genomes/c.elegans.