public static List <SNPItem> ReadFromFile(string fileName)
{
var result = new List <SNPItem>();
using (var sr = new StreamReader(fileName))
{
var line = sr.ReadLine();
var hasFlip = line.Contains("_AlleleNeedFlip");
var platforms = line.Split('\t').ToList().ConvertAll(m => m.StringBefore("_Allele"));
var step = hasFlip ? 3 : 2;
while ((line = sr.ReadLine()) != null)
{
var parts = line.Split('\t');
//sw.WriteLine("Dbsnp_ID\tCHROM\tPOS\tDbsnp_RefAllele\tDbsnp_AltAllele\tDbsnp_IsReversed\t1000G_REF\t1000G_ALT\t1000G_MAF\tDataset\t{0}",
var item = new SNPItem()
{
Name = parts[0],
Chrom = int.Parse(parts[1]),
Position = int.Parse(parts[2]),
Allele1 = parts[3][0],
Allele2 = parts[4],
DbsnpIsReversed = bool.Parse(parts[5]),
G1000Allele1 = parts[6][0],
G1000Allele2 = parts[7][0],
G1000Allele2Frequency = double.Parse(parts[8]),
Dataset = parts[9]
};
for (int i = 10; i < parts.Length; i += step)
{
if (string.IsNullOrEmpty(parts[i]))
{
continue;
}
var alleles = parts[i].Split(':');
var aitem = new Alleles()
{
Allele1 = alleles[0],
Allele2 = alleles[1]
};
aitem.Allele2Frequency = double.Parse(parts[i + 1]);
if (hasFlip)
{
aitem.NeedFlip = bool.Parse(parts[i + 2]);
}
item.Platforms[platforms[i]] = aitem;
}
result.Add(item);
}
}
return(result);
}
public override IEnumerable <string> Process()
{
var targetSNPs = SNPItem.ReadFromFile(_options.TargetSnpFile);
using (var sw = new StreamWriter(_options.OutputFile))
using (var swInfo = new StreamWriter(_options.OutputFile + ".info"))
{
swInfo.WriteLine("MarkerId\tIsImputed");
foreach (var file in _options.InputFiles)
{
int chromosome = DetectChromosome(file);
Progress.SetMessage("Chromosome {0} : {1}", chromosome, file);
var locusMap = targetSNPs.Where(m => m.Chrom == chromosome).ToDictionary(m => m.Position.ToString());
using (var sr = new StreamReader(file))
{
string line;
SNPItem item;
while ((line = sr.ReadLine()) != null)
{
string[] parts = line.Take(delimiter, 3);
bool isImputed = parts[0].Equals("---");
if (isImputed)
{
if (locusMap.TryGetValue(parts[2], out item))
{
var name = string.IsNullOrEmpty(item.Name) ? parts[1] : item.Name;
var markerid = string.IsNullOrEmpty(item.Dataset) ? name : item.Dataset + ":" + name;
sw.WriteLine("{0} {1}{2}",
item.Chrom,
markerid,
line.StringAfter(parts[1]));
swInfo.WriteLine("{0}\t{1}", markerid, isImputed);
}
}
else
{
sw.WriteLine(line);
swInfo.WriteLine("{0}\t{1}", parts[1], isImputed);
}
}
}
}
}
return(new[] { _options.OutputFile });
}
private static void DoFillAllele2FrequencyFrom1000Gome(this IEnumerable <SNPItem> snpItems, string g1000VcfFile, Func <SNPItem, string> keyFunc, IProgressCallback progress = null)
{
if (progress == null)
{
progress = new ConsoleProgressCallback();
}
var dic = snpItems.ToDictionary(m => keyFunc(m));
progress.SetMessage("Filling MAF from {0} ...", g1000VcfFile);
using (var sr = new StreamReader(g1000VcfFile))
{
progress.SetRange(0, sr.BaseStream.Length);
string line;
while ((line = sr.ReadLine()) != null)
{
if (!line.StartsWith("##"))
{
break;
}
}
int linecount = 0;
while ((line = sr.ReadLine()) != null)
{
linecount++;
if (linecount % 10000 == 0)
{
progress.SetPosition(sr.GetCharpos());
}
var parts = line.Split('\t');
var snp = new SNPItem()
{
Chrom = HumanChromosomeToInt(parts[0]),
Position = int.Parse(parts[1]),
Name = parts[2]
};
SNPItem loc;
if (!dic.TryGetValue(keyFunc(snp), out loc))
{
continue;
}
loc.G1000Allele1 = parts[3][0];
var allele2 = parts[4].Split(',');
var frequencies = parts[7].StringAfter("AF=").StringBefore(";").Split(',');
bool bFound = false;
for (int i = 0; i < allele2.Length; i++)
{
if (allele2[i].Length != 1)
{
continue;
}
loc.G1000Allele2 = allele2[i][0];
if (loc.IsSourceAllelesMatchedWithG1000())
{
loc.G1000Allele2Frequency = double.Parse(frequencies[i]);
bFound = true;
break;
}
}
if (!bFound)
{
loc.G1000Allele1 = ' ';
loc.G1000Allele2 = ' ';
loc.G1000Allele2Frequency = 0.0;
}
}
}
progress.SetMessage("Filling MAF finished.");
}
public static List<SNPItem> ReadFromFile(string fileName)
{
var result = new List<SNPItem>();
using (var sr = new StreamReader(fileName))
{
var line = sr.ReadLine();
var hasFlip = line.Contains("_AlleleNeedFlip");
var platforms = line.Split('\t').ToList().ConvertAll(m => m.StringBefore("_Allele"));
var step = hasFlip ? 3 : 2;
while ((line = sr.ReadLine()) != null)
{
var parts = line.Split('\t');
//sw.WriteLine("Dbsnp_ID\tCHROM\tPOS\tDbsnp_RefAllele\tDbsnp_AltAllele\tDbsnp_IsReversed\t1000G_REF\t1000G_ALT\t1000G_MAF\tDataset\t{0}",
var item = new SNPItem()
{
Name = parts[0],
Chrom = int.Parse(parts[1]),
Position = int.Parse(parts[2]),
Allele1 = parts[3][0],
Allele2 = parts[4],
DbsnpIsReversed = bool.Parse(parts[5]),
G1000Allele1 = parts[6][0],
G1000Allele2 = parts[7][0],
G1000Allele2Frequency = double.Parse(parts[8]),
Dataset = parts[9]
};
for (int i = 10; i < parts.Length; i += step)
{
if (string.IsNullOrEmpty(parts[i]))
{
continue;
}
var alleles = parts[i].Split(':');
var aitem = new Alleles()
{
Allele1 = alleles[0],
Allele2 = alleles[1]
};
aitem.Allele2Frequency = double.Parse(parts[i + 1]);
if (hasFlip)
{
aitem.NeedFlip = bool.Parse(parts[i + 2]);
}
item.Platforms[platforms[i]] = aitem;
}
result.Add(item);
}
}
return result;
}
private static void DoFillAllele2FrequencyFrom1000Gome(this IEnumerable<SNPItem> snpItems, string g1000VcfFile, Func<SNPItem, string> keyFunc, IProgressCallback progress = null)
{
if (progress == null)
{
progress = new ConsoleProgressCallback();
}
var dic = snpItems.ToDictionary(m => keyFunc(m));
progress.SetMessage("Filling MAF from {0} ...", g1000VcfFile);
using (var sr = new StreamReader(g1000VcfFile))
{
progress.SetRange(0, sr.BaseStream.Length);
string line;
while ((line = sr.ReadLine()) != null)
{
if (!line.StartsWith("##"))
{
break;
}
}
int linecount = 0;
while ((line = sr.ReadLine()) != null)
{
linecount++;
if (linecount % 10000 == 0)
{
progress.SetPosition(sr.GetCharpos());
}
var parts = line.Split('\t');
var snp = new SNPItem()
{
Chrom = HumanChromosomeToInt(parts[0]),
Position = int.Parse(parts[1]),
Name = parts[2]
};
SNPItem loc;
if (!dic.TryGetValue(keyFunc(snp), out loc))
{
continue;
}
loc.G1000Allele1 = parts[3][0];
var allele2 = parts[4].Split(',');
var frequencies = parts[7].StringAfter("AF=").StringBefore(";").Split(',');
bool bFound = false;
for (int i = 0; i < allele2.Length; i++)
{
if (allele2[i].Length != 1)
{
continue;
}
loc.G1000Allele2 = allele2[i][0];
if (loc.IsSourceAllelesMatchedWithG1000())
{
loc.G1000Allele2Frequency = double.Parse(frequencies[i]);
bFound = true;
break;
}
}
if (!bFound)
{
loc.G1000Allele1 = ' ';
loc.G1000Allele2 = ' ';
loc.G1000Allele2Frequency = 0.0;
}
}
}
progress.SetMessage("Filling MAF finished.");
}
