Appeal No. 2006-1569 Page 2
Application No. 10/159,997
from which alternative splicing information is commonly derived, researchers often
restrict their analyses to the simple splicing modes. . . . A more complete understanding
of alternative splicing requires an unbiased consideration of all alternative mRNA
isoforms.” Specification, page 1.
The specification discloses “a protocol to systematically identify alternative
mRNA isoforms of known human genes.” Page 3. The specification describes “an
exemplary implementation of this protocol” as follows:
[W]e map mRNAs from the RefSeq database to contig sequences from
the NCBI human genome, requiring that an mRNA align to genomic
sequence over the full length of the coding sequence, without gaps in the
exons. We further require 98% identity between the sequences, favoring
RefSeq sequence in case of nucleotide mismatch. When multiple RefSeq
mRNAs align to the same region of genomic sequence, we use only the
mRNA containing the largest number of exons.
Page 4. This part of the protocol “identif[ies] target gene sequences.” Page 3,
lines 13-21. The subsequent steps are described as follows:
To detect alternate isoforms, we align EST sequences from dbEST to the
genomic sequence and use TAP [transcript assembly protocol;
specification, page 4, line 7] to infer alternate mRNA splice forms from
these alignments. Since we use known genes, the reading frame of the
primary mRNA isoforms (i.e., the RefSeq mRNAs) is known. So that the
reading frame can be determined for all EST-suggested alternate
isoforms, we restrict our set to cases in which the 5’ end of the EST
sequences align to coding sequences of the RefSeq mRNA. We also
exclude cases of intron retention, as these are indistinguishable from
incompletely-processed transcripts, a common dbEST contaminant. We
have higher confidence in splicing events with coverage by multiple ESTs.
Page 4.
Page: Previous 1 2 3 4 5 6 7 8 9 10 11 Next
Last modified: November 3, 2007