/// <summary>
/// This method is called for every Lambda invocation. This method takes in an S3 event object and can be used
/// to respond to S3 notifications.
/// </summary>
/// <param name="evnt"></param>
/// <param name="context"></param>
/// <returns></returns>
public async Task BPMToBIMLambdaFunctionHandler(S3Event evnt, ILambdaContext context)
{
JObject preference = new Preference().preference;
var s3Event = evnt.Records?[0].S3;
if (s3Event == null)
{
return;
}
try
{
var bpm = new BeadPoolManifest(new BinaryReader(readStreamFromS3ToMemory(bucketName: s3Event.Bucket.Name,
keyName: s3Event.Object.Key,
bucketRegion: preference["S3"]["BPM"]["Region"].ToString()).Result));
var ms = new MemoryStream();
var sw = new StreamWriter(ms);
for (int idx = 0; idx < bpm.num_loci; ++idx)
{
List <string> bimEntry = new List <string> {
bpm.chroms[idx], bpm.names[idx], "0", bpm.map_infos[idx].ToString()
};
List <string> alleles = bpm.ref_strands[idx] == 1 ? bpm.snps[idx].Replace("[", "").Replace("]", "").Split('/').ToList() :
bpm.snps[idx].Replace("[", "").Replace("]", "").Split('/').Select(x => Utils.getComplement(x[0]).ToString()).ToList();
bimEntry.AddRange(alleles);
sw.WriteLine(String.Join(' ', bimEntry));
sw.Flush();
}
writeStreamFromStreamToS3(bucketName: s3Event.Bucket.Name,
keyName: s3Event.Object.Key.Replace("bpm", "bim"),
bucketRegion: preference["S3"]["BPM"]["Region"].ToString(),
stream: ms).Wait();
}
catch (Exception e)
{
context.Logger.LogLine(e.Message);
context.Logger.LogLine(e.StackTrace);
throw;
}
}
/*
* Method for amending binary(BPM) manifest file.
*
* Attributes:
* bpm (BeadPoolManifest): original manifest file
* amend_file: amend info file (csv format, no header)
* amend (list({'from': SNP name in original manifest,
* 'to': SNP name in new manifest,
* 'ref_strands': ref_strands in new manifest,
* 'snps': snps in new manifest}))
*
* names (list(strings): Names of loci from manifest
* snps (list(strings): SNP values of loci from manifest
*
* chroms (list(strings): Chromosome values for loci
* map_infos (list(int)): Map infor values for loci
* addresses(list(int)): AddressA IDs of loci from manifest
* normalization_lookups(list(int)) : Normalization lookups from manifest.This indexes into
* list of normalization transforms read from GTC file
* ref_strands (list(int)): Reference strand annotation for loci(see RefStrand class)
* source_strands(list(int)) : Source strand annotations for loci(see SourceStrand class)
* num_loci(int) : Number of loci in manifest
* manifest_name(string): Name of manifest
* control_config(string): Control description from manifest
*/
public static BeadPoolManifest Amend(BeadPoolManifest bpm, List <BeadPoolManifestAmendEntry> amendEntries)
{
for (int idx = 0; idx < bpm.num_loci; ++idx)
{
foreach (BeadPoolManifestAmendEntry amendEntry in amendEntries)
{
if (amendEntry.originalSNPNameOnBPM == bpm.names[idx])
{
// Console.WriteLine("Index: " + idx + ", locus_name: " + bpm.locus_entries[idx].name + ", name: " + x["to"] + ", Strand: " + x["ref_strands"] + ", snps: " + x["snps"]);
bpm.names[idx] = amendEntry.newSNPNameOnBPM;
// bpm.ref_strands[idx] = Convert.ToInt32(x["ref_strands"]);
bpm.locus_entries[idx].ref_strand = amendEntry.newStrandOnBPM;
// bpm.snps[idx] = x["snps"];
bpm.locus_entries[idx].snp = amendEntry.newAlleleOnBPM;
bpm.locus_entries[idx].name = amendEntry.newSNPNameOnBPM;
}
}
}
return(bpm);
}
public static Plink.Plink ConvertGTCToPlink(List <GenotypeCalls> gtcs, BeadPoolManifest bpm = null)
{
Dictionary <char, byte> sex_map = new Dictionary <char, byte>
{
{ 'F', 0x02 },
{ 'M', 0x01 },
{ 'U', 0x00 }
};
List <string> FIDs = gtcs.AsParallel().Select(gtc => gtc.get_sample_name()).ToList();
List <string> parent = gtcs.AsParallel().Select(gtc => "0").ToList();
List <byte> Sexs = gtcs.AsParallel().Select(gtc => sex_map[gtc.get_gender()]).ToList();
List <byte> Phenotypes = gtcs.AsParallel().Select(gtc => (byte)1).ToList();
int num_snps = gtcs[0].get_genotypes().Length;
List <byte[]> genotypesGTC = gtcs.AsParallel().Select(gtc => gtc.get_genotypes()).ToList();
List <byte[]> genotypes = new List <byte[]>();
for (int snpIdx = 0; snpIdx < num_snps; ++snpIdx)
{
genotypes.Add(new byte[gtcs.Count]);
for (int gtcIdx = 0; gtcIdx < gtcs.Count; ++gtcIdx)
{
genotypes[snpIdx][gtcIdx] = genotypesGTC[gtcIdx][snpIdx];
}
}
Console.WriteLine("GTC data loaded; Converting to Plink");
return(new Plink.Plink(FIDs: FIDs,
IIDs: FIDs,
Fathers: parent,
Mothers: parent,
Sexs: Sexs,
Phenotypes: Phenotypes,
Chromosomes: bpm == null ? null : bpm.chroms,
snpNames: bpm == null ? null : bpm.names,
cMs: bpm == null ? null : Enumerable.Range(0, bpm.num_loci).Select(x => 0).ToList(),
Positions: bpm == null ? null : bpm.map_infos,
alleles: bpm == null ? null : bpm.snps.AsParallel().Select(allele => allele.Replace("[", "").Replace("]", "").Split('/').ToList()).ToList(),
genotypes: genotypes,
strands: bpm == null ? null : bpm.ref_strands.Select(x => Convert.ToByte(x)).ToList()));
}
public static MemoryStreamNotDispose beadPoolManifestWriter(BeadPoolManifest bpm)
{
MemoryStreamNotDispose fs = new MemoryStreamNotDispose();
using (BinaryWriter bw = new BinaryWriter(fs))
{
for (int locus_idx = 0; locus_idx < bpm.num_loci; ++locus_idx)
{
bpm.normalization_ids[locus_idx] -= 100 * bpm.assay_types[locus_idx];
}
bw.Write(System.Text.Encoding.ASCII.GetBytes("BPM"));
bw.Write((byte)1);
bw.Write(bpm.version);
bw.Write(bpm.manifest_name);
bw.Write(bpm.control_config);
bw.Write(bpm.num_loci);
bw.Write(bpm.after_num_loci);
for (int idx = 0; idx < bpm.num_loci; ++idx)
{
bw.Write(bpm.names[idx]);
}
for (int idx = 0; idx < bpm.num_loci; ++idx)
{
bw.Write((byte)bpm.normalization_ids[idx]);
}
for (int idx = 0; idx < bpm.num_loci; ++idx)
{
LocusEntryWriter.Writer(bw, bpm.locus_entries[idx]);
}
}
return(fs);
}
