public void TrimPrefix()
{
// RSS/R
var hn = new HgvsProteinNomenclature.HgvsNotation("RSS", "R", "bob", 100, 102)
{
Type = ProteinChange.Deletion
};
AminoAcids.RemovePrefixAndSuffix(hn);
const string expectedReference = "SS";
Assert.Equal(expectedReference, hn.ReferenceAminoAcids);
const string expectedAlternate = null;
Assert.Equal(expectedAlternate, hn.AlternateAminoAcids);
const int expectedStart = 101;
Assert.Equal(expectedStart, hn.Start);
const int expectedEnd = 102;
Assert.Equal(expectedEnd, hn.End);
}
public void TrimBothPrefixAndSuffix()
{
// RT/RMLMLT
var hn = new HgvsProteinNomenclature.HgvsNotation("RT", "RMLMLT", "bob", 100, 101)
{
Type = ProteinChange.Insertion
};
AminoAcids.RemovePrefixAndSuffix(hn);
const string expectedReference = null;
Assert.Equal(expectedReference, hn.ReferenceAminoAcids);
const string expectedAlternate = "MLML";
Assert.Equal(expectedAlternate, hn.AlternateAminoAcids);
const int expectedStart = 101;
Assert.Equal(expectedStart, hn.Start);
const int expectedEnd = 100;
Assert.Equal(expectedEnd, hn.End);
}
/// <summary>
/// given a common amino acid prefix, remove the common amino acids (insertion)
/// returns true if the alleles were modified, false otherwise
/// </summary>
internal static void RemovePrefixAndSuffix(HgvsProteinNomenclature.HgvsNotation hn)
{
// nothing to do if we have a pure insertion or deletion
if (hn.ReferenceAminoAcids == null || hn.AlternateAminoAcids == null)
{
return;
}
// skip this if the amino acids are already the same
if (hn.ReferenceAminoAcids == hn.AlternateAminoAcids)
{
return;
}
// calculate how many shared amino acids we have from the beginning of each amino acid
var numSharedPrefixPos = 0;
var isClipped = false;
var refLen = hn.ReferenceAminoAcids.Length;
var altLen = hn.AlternateAminoAcids.Length;
var minLength = Math.Min(refLen, altLen);
for (var pos = 0; pos < minLength; pos++, numSharedPrefixPos++, hn.Start++)
{
if (hn.ReferenceAminoAcids[pos] != hn.AlternateAminoAcids[pos])
{
break;
}
refLen--;
altLen--;
isClipped = true;
}
// calculate how many shared amino acids we have from the end of each amino acid
minLength = Math.Min(refLen, altLen);
for (var pos = 0; pos < minLength; pos++, hn.End--)
{
var refPos = hn.ReferenceAminoAcids.Length - pos - 1;
var altPos = hn.AlternateAminoAcids.Length - pos - 1;
if (hn.ReferenceAminoAcids[refPos] != hn.AlternateAminoAcids[altPos])
{
break;
}
refLen--;
altLen--;
isClipped = true;
}
// clip the amino acid alleles
if (isClipped)
{
hn.SetReferenceAminoAcids(refLen == 0 ? null : hn.ReferenceAminoAcids.Substring(numSharedPrefixPos, refLen));
hn.SetAlternateAminoAcids(altLen == 0 ? null : hn.AlternateAminoAcids.Substring(numSharedPrefixPos, altLen));
}
}
// ReSharper disable once InconsistentNaming
public void Shift3PrimeSS()
{
// given a SS/- deletion in RS[SS]SSS, we want to move to: RSSSS[SS]
const string transcriptPeptides = "RSSSSSS";
var hn = new HgvsProteinNomenclature.HgvsNotation("SS", null, "bob", 3, 4)
{
Type = ProteinChange.Deletion
};
AminoAcids.Rotate3Prime(hn, transcriptPeptides);
Assert.Equal(6, hn.Start);
Assert.Equal(7, hn.End);
}
// ReSharper disable once InconsistentNaming
public void Shift3PrimeSTM()
{
// given a STM/- deletion in R[STM]STMP, we want to move to: RSTM[STM]P
const string transcriptPeptides = "RSTMSTMP";
var hn = new HgvsProteinNomenclature.HgvsNotation("STM", null, "bob", 2, 4)
{
Type = ProteinChange.Deletion
};
AminoAcids.Rotate3Prime(hn, transcriptPeptides);
Assert.Equal(5, hn.Start);
Assert.Equal(7, hn.End);
}
internal static void Rotate3Prime(HgvsProteinNomenclature.HgvsNotation hn, string peptides)
{
if (hn.Type != ProteinChange.Deletion &&
hn.Type != ProteinChange.Duplication &&
hn.Type != ProteinChange.Insertion
)
{
return;
}
// for insertion, the reference bases will be empty string. The shift should happen on the alternate allele
var rotatingPeptides = hn.Type == ProteinChange.Insertion ? hn.AlternateAminoAcids : hn.ReferenceAminoAcids;
var numBases = rotatingPeptides.Length;
var downstreamPeptides = peptides.Length >= hn.End ? peptides.Substring(hn.End): null;
if (downstreamPeptides == null)
{
return;
}
var combinedSequence = rotatingPeptides + downstreamPeptides;
int shiftStart, shiftEnd;
var hasShifted = false;
for (shiftStart = 0, shiftEnd = numBases; shiftEnd < combinedSequence.Length; shiftStart++, shiftEnd++)
{
if (combinedSequence[shiftStart] != combinedSequence[shiftEnd])
{
break;
}
hn.Start++;
hasShifted = true;
}
if (hasShifted)
{
rotatingPeptides = combinedSequence.Substring(shiftStart, numBases);
}
if (hn.Type == ProteinChange.Insertion)
{
hn.AlternateAminoAcids = rotatingPeptides;
}
else
{
hn.ReferenceAminoAcids = rotatingPeptides;
}
hn.End = hn.Type == ProteinChange.Insertion ? hn.Start - 1 : hn.Start + numBases - 1;
if (hn.Type != ProteinChange.Insertion || !hasShifted)
{
return;
}
var newUpstreamSeq = combinedSequence.Substring(0, shiftStart);
if (newUpstreamSeq.EndsWith(rotatingPeptides))
{
hn.Type = ProteinChange.Duplication;
// We are not sure why we have to take 1 off numBases. But that is what VEP is doing.
// var incrementLength = numBases;
// hn.Start -= incrementLength;
hn.End = hn.Start + numBases - 1;
hn.ReferenceAminoAcids = hn.AlternateAminoAcids;
}
}
