+ protected static boolean myShortSequenceNucleotideProportionCount(
+ long ntCount, long allCount)
+ {
+ /**
+ * this method is valid only for NUCLEOTIDE_COUNT_VERY_SHORT_SEQUENCE <=
+ * allCount <= NUCLEOTIDE_COUNT_SHORT_SEQUENCE
+ */
+ // the following is a simplified integer version of:
+ //
+ // a := allCount # the number of bases in the sequence
+ // n : = ntCount # the number of definite nucleotide bases
+ // vs := NUCLEOTIDE_COUNT_VERY_SHORT_SEQUENCE
+ // s := NUCLEOTIDE_COUNT_SHORT_SEQUENCE
+ // lp := NUCLEOTIDE_COUNT_LOWER_PERCENT
+ // vsp := 1 - (1/a) # this is the proportion of required definite
+ // nucleotides
+ // # in a VERY_SHORT Sequence (4 bases).
+ // # This should be equivalent to all but one base in the sequence.
+ // p := (a - vs)/(s - vs) # proportion of the way between
+ // # VERY_SHORT and SHORT thresholds.
+ // tp := vsp + p * (lp/100 - vsp) # the proportion of definite nucleotides
+ // # required for this length of sequence.
+ // minNt := tp * a # the minimum number of definite nucleotide bases
+ // # required for this length of sequence.
+ //
+ // We are then essentially returning:
+ // # ntCount >= 55% of allCount and the rest are all nucleotide ambiguity:
+ // ntCount >= tp * allCount && nCount + ntaCount == aaCount
+ // but without going into float/double land
+ long LHS = 100 * allCount
+ * (NUCLEOTIDE_COUNT_SHORT_SEQUENCE
+ - NUCLEOTIDE_COUNT_VERY_SHORT_SEQUENCE)
+ * (ntCount - allCount + 1);
+ long RHS = allCount * (allCount - NUCLEOTIDE_COUNT_VERY_SHORT_SEQUENCE)
+ * (allCount * NUCLEOTIDE_COUNT_PERCENT - 100 * allCount + 100);
+ return LHS >= RHS;