Description

Description: Results of RT-PCR testing of novel genes in $organism done in collaboration with Phil Green's lab (U. Washington and HHMI). This track shows predictions that are currently being tested.

Methods

Predicted gene fragments from the ExoniphyGenes track were tested. To be selected, predictions were required not to overlap RefSeq or Vega genes, UCSC Known Genes linked to SwissProt entries, or MGC A- or B-list genes. They also had to contain at least one intron, could not fall within introns of, and on the same strand as, known genes, and could not be in recent segmental duplications in the human genome (with >95% identity over >1000bp to other regions of the genome); moreover, they could not overlap C-list predictions tested or in the pipeline to be tested by the Brent lab or the Green lab. Predicted human genes and orthologous predictions in mouse were tested in parallel (only the human results are shown here). When choosing predictions to be tested, priority was given to those with little or no cDNA support in human or mouse; however, some predictions with significant cDNA evidence in one or both species have been tested.

Sequence reads were assembled using Phrap (if possible) and then aligned to the genome using BLAT. Currently, a test is considered a success if the sequenced product (either assembled or the raw reads, if assembly is not possible) matches the genome with at least 95% identity and 30% coverage in the same location as the original prediction, and the alignment confirms that at least one GT-AG intron has been spliced out. (A low coverage threshold is used because of a tendency for poor sequence quality near the ends of the reads.) If any of these criteria are not met, the test is considered a failure.

Credits

Phil Green, Colleen Davis, and Brent Ewing at the Green Lab (http://www.phrap.org).