Documentation

ValueObject

Table of Contents

Classes

CodonUsageTable
Holds raw codon observation counts from a reference gene set (typically the highly expressed genes of a host organism), keyed by codon in DNA notation. This is deliberately just a validated data carrier : it knows nothing about which codons are synonymous with each other, since that grouping comes from the genetic code, not from this table, and is CodonAdaptationIndexCalculator's concern.
CpGIsland
Coordinates are 1-based inclusive, the convention the rest of this project uses. gcContent and observedToExpectedRatio are computed over the island's own full span - once it is known, not inherited from whichever single scanning window first triggered its detection - so they describe the island actually reported, not an arbitrary window inside it.
GeneticCodeTable
Numbering matches NCBI's own "Genetic Codes" table IDs, so a caller already familiar with that numbering (or reading it off a GenBank /transl_table qualifier) can use it directly. The tables are those of NCBI's gc.prt (version 4.6) whose stop codons are the same in every context, which Biopython's CodonTable agrees with on all 64 codons. Left out are 27 (Karyorelict), 28 (Condylostoma) and 31 (Blastocrithidia) : their UGA, and for the last two UAA and UAG, is a stop or an amino acid according to its position, which a table giving one reading per codon cannot say.
OpenReadingFrame
frame follows the standard six-frame convention : +1/+2/+3 for the forward strand (the reading frame's zero-based offset plus one), -1/-2/-3 for the reverse strand (the same, on the reverse complement). start/end are always given as ascending 1-based inclusive positions in the ORIGINAL, forward-strand sequence - even for a reverse-strand ORF, the same convention GenBank's own complement(start..end) location uses, so a negative frame's coordinates stay directly comparable to a positive one's. end includes the stop codon's three bases when hasStopCodon() is true ; getPeptide() itself never includes the "*" stop symbol.
SkewResult
A skew with an undefined denominator (e.g. GC-skew over a window with no G or C at all) is 0.0 rather than thrown or NAN - a deliberate, documented choice (SkewCalculator's own docblock) so a sliding window never has to special-case an all-A/T or otherwise degenerate window.
On this page

Search results