def testKSpectrum(self):
spec = nc.features.kSpectrum(5)
testSeq = nc.sequence('', 'GAGAGAGAGT')
kmerFreq = {
k: v
for k, v in zip([f.kmer for f in spec], spec.getAll(testSeq))
}
self.assertEqual(kmerFreq['AGAGT'], 100.)
self.assertEqual(kmerFreq['ACTCT'], 100.)
self.assertEqual(kmerFreq['AGAGA'], 200.)
self.assertEqual(kmerFreq['TCTCT'], 200.)
self.assertEqual(kmerFreq['GAGAG'], 300.)
self.assertEqual(kmerFreq['CTCTC'], 300.)
self.assertEqual(
sum(kmerFreq[kmer] for kmer in kmerFreq.keys() if kmer not in
['AGAGT', 'ACTCT', 'AGAGA', 'TCTCT', 'GAGAG', 'CTCTC']), 0.)
# Ensure that application to the reverse complement yields an identical spectrum
specmotifs = set(k
for k, f in zip([f.kmer
for f in spec], spec.getAll(testSeq))
if f > 0)
specmotifsRC = set(k for k, f in zip(
[f.kmer for f in spec],
spec.getAll(
nc.sequence(
'', nc.getReverseComplementaryDNASequence(testSeq.seq))))
if f > 0)
self.assertEqual(specmotifs, specmotifsRC)
def testKSpectrumMM(self):
spec = nc.features.kSpectrumMM(5)
testSeq = nc.sequence('', 'GAGAGAGAGT')
motifs = set([ testSeq.seq[i:i+5] for i in range(len(testSeq)-4) ])
motifs = motifs | set( nc.getReverseComplementaryDNASequence(m) for m in motifs )
def getMM(m):
return set(
m[:i] + mut + m[i+1:]
for i in range(len(m))
for mut in [ 'A', 'C', 'G', 'T' ]
)
motifs = motifs | set( mm for m in motifs for mm in getMM(m) )
# Ensure that only the main motif and mismatches are registered
specmotifs = set(
k
for k, f in zip(
[ f.kmer for f in spec ],
spec.getAll(testSeq)
)
if f > 0
)
self.assertEqual(specmotifs, motifs)
# Ensure that application to the reverse complement yields an identical spectrum
specmotifsRC = set(
k
for k, f in zip(
[ f.kmer for f in spec ],
spec.getAll(nc.sequence('',
nc.getReverseComplementaryDNASequence(testSeq.seq)))
)
if f > 0
)
self.assertEqual(specmotifs, specmotifsRC)
def prep():
global genome
global rsA, rsB
global PcG
global gwWin
global PcGTargets
global PRESeq
global MC
global testSeqs
system('rm -rf ./temp')
system('mkdir ./temp')
genome = nc.streamFASTAGZ('tutorial/DmelR5.fasta.gz',
restrictToSequences = [ '2L', '2R', '3L', '3R', '4', 'X' ])
#
rsA = nc.regions( 'A', [
nc.region('X', 20, 50), nc.region('X', 80, 100), nc.region('X', 150, 300), nc.region('X', 305, 400), nc.region('X', 500, 600),
nc.region('Y', 40, 100), nc.region('Y', 120, 200)
] )
#
rsB = nc.regions( 'B', [
nc.region('X', 30, 40), nc.region('X', 90, 120), nc.region('X', 140, 150), nc.region('X', 300, 310),
nc.region('Y', 40, 100), nc.region('Y', 130, 300), nc.region('Y', 600, 700)
] )
#
# Prepare data set
#
PcG = nc.biomarkers('PcG', [
nc.loadGFFGZ('tutorial/Pc.gff3.gz').deltaResize(1000).rename('Pc'),
nc.loadGFFGZ('tutorial/Psc.gff3.gz').deltaResize(1000).rename('Psc'),
nc.loadGFFGZ('tutorial/dRING.gff3.gz').deltaResize(1000).rename('dRING'),
nc.loadGFFGZ('tutorial/H3K27me3.gff3.gz').rename('H3K27me3'),
])
#
gwWin = nc.getSequenceWindowRegions(
genome,
windowSize = 1000, windowStep = 100)
#
PcGTargets = PcG.HBMEs(gwWin, threshold = 4)
#
PRESeq = PcGTargets.recenter(3000).extract(genome)
random.shuffle(PRESeq.sequences)
#
MC = nc.MarkovChain(trainingSequences = genome, degree = 4, pseudoCounts = 1, addReverseComplements = True)
#
testSeqs = nc.sequences('Test', [
nc.sequence('X', ''.join( random.choice(['A', 'C', 'G', 'T'])
for _ in range(800) )),
nc.sequence('Y', ''.join( random.choice(['A', 'C', 'G', 'T'])
for _ in range(1000) )),
])
import random
import gnocis as nc
genome = nc.streamFASTAGZ(
'tutorial/DmelR5.fasta.gz',
restrictToSequences=['2L', '2R', '3L', '3R', '4', 'X'])
testSeqs = nc.sequences('Test', [
nc.sequence(
'X', ''.join(random.choice(['A', 'C', 'G', 'T']) for _ in range(800))),
nc.sequence(
'Y', ''.join(random.choice(['A', 'C', 'G', 'T'])
for _ in range(1000))),
])