Glossary

Information Theory-Based Terms

:

• R_i: The individual information for a splice site defined as the dot product of an information theory-based weight matrix and the unitary vector of a particular splice site sequence. This value represents binding affinity and is measured in bits. The minimum fold change in binding affinity resulting from the mutation is an exponential function based on ΔR_i, or ≥ 2ΔR_i. Therefore, a 3 bit decrease would correspond to a ≥ 8-fold decrease in binding affinity.
• R_i,min: the minimum strength of functional splice sites, based on 103 mutations with functional validation, including 57 natural and 46 cryptic site activating mutations. Computed to be 1.6 bits (Rogan et al., 2003)
• R_i before mutation: Also known as R_i,initial, it is the predicted strength of the wildtype splice site.
• R_i after mutation: Also known as R_i,final, it is the predicted strength of the splice site in the presence of the mutation of interest.
• ΔR_i: The total change in information content (R_i,final - R_i,initial)
• Site Type “NATURALSITE”: The mutation of interest is weakening the constitutive splice site. Mutations that affect these sites can often lead to increased exon skipping, total intron inclusion, and use of a nearby alternate splice site (cryptic site)
• Site Type “CRYPTICSITE”: The mutation of interest creates or strengthens a non-constitutive splice site (a site not recognized in normal splicing). A variant strengthening a cryptic site may simultaneously weaken a natural site, which is indicated when searching a variant in ValidSpliceMut.
• Splice Type “DONOR” and “ACCEPTOR”: The splice sites found at the 5’ (donor) and 3’ (acceptor) ends of an intron. When two sites are referred to as being in the same phase, they are either both donors or both acceptors.
• Inactivated / Abolished Splicing: Mutation weakens the natural splice site to a degree where it is no longer expected to be recognized by the splicing machinery.
• Leaky Splicing: Mutation weakens but does not completely inactivate the constitutive splice site. Often leads to two (or more) distinct mRNA isoforms, including a diminished amount of wildtype splicing. Natural site mutations where the R_i,final ≥ R_i,min (1.6 bits) are predicted to be leaky.

Veridical Terms (text from Viner et al. 2014)

:

• Junction-spanning reads: “Contain DNA sequences from two adjacent exons or are reads that extend into the intron. Directly show the intronic sequence removed by the spliceosome."
• Read-abundance reads: “Reads you would expect to see based upon which sequences are predicted to be found in the transcript versus those reads that are expected to be spliced out.
• Cryptic site use: “Reads using abnormal exon boundaries in transcripts using non-constitutive, proximate sequences, extending or truncating the exon.”
• Exon skipping: “Reads which contain connecting segments of reads split across an exon.”
• Total intron inclusion: “Reads present in an intronic region of the genome, including all reads classified within the "intron inclusion with mutation" category.”
• Intron inclusion with mutation: “Intron inclusion reads, where the read contains the predicted mutation.”
• Cryptic site use/anti cryptic site use: “Equivalent to: cryptic corroborating / non-cryptic corroborating reads”
• Cryptic corroborating reads: "Reads occurring within the expected region for cryptic splicing to occurring (i.e. spliced-in regions). This region is between the variant splice site location and the adjacent (direction dependent) splice junction."
• Non-cryptic corroborating reads: "Also termed "anti-cryptic" reads. Not within the expected region, but still within the portion that would be expected to be spliced out, had cryptic splicing occurred."