R interface to SeqLib C++ package
devtools::install_github('mskilab/RSeqLib')## instantiate the BWA object around a character vector custom reference sequence
bwa = BWA(seq = c("CACTAGCTAGCTACGCGGGGGCGCGCGCGCGCGAAAAACACTTTCACAG"))
## align two sequences to this reference using "[" operator
## returned result is a granges on the above sequence
bwa[c("CACTAGCTAGCTACGCGGGGGCGCG", "CACTAGCTAGCTACGCGCGAAAAACACTTTCACAG")]GRanges object with 2 ranges and 12 metadata columns:
seqnames ranges strand | qname flag mapq
<Rle> <IRanges> <Rle> | <character> <character> <character>
[1] 1 [ 0, 24] + | myquery 0 36
[2] 1 [27, 48] + | myquery 0 18
cigar rnext pnext tlen
<character> <character> <character> <character>
[1] 25M 0 -1 0
[2] 13S22M 0 -1 0
seq qual AS DD
<character> <character> <integer> <integer>
[1] CACTAGCTAGCTACGCGGGGGCGCG * 25 0
[2] CACTAGCTAGCTACGCGCGAAAAACACTTTCACAG * 22 0
qwidth
<integer>
[1] 25
[2] 35
-------
seqinfo: 1 sequence from an unspecified genome; no seqlengths
## you can make a reference genome with a multiple named contigs
## in this case we make them identical
bwa = BWA(seq = c(
allele1 = "CACTAGCTAGCTACGCGGGGGCGCGCGCGCGCGAAAAACACTTTCACAG",
allele2 = "CACTAGCTAGCTACGCGGGGGCGCGCGCGCGCGAAAAACACTTTCACAG"))
## notice that now the seqnames of alignments are "allele1" and "allele2"
## and we get multiple alignments per query, each with mapq0, as expected
bwa[c("CACTAGCTAGCTACGCGGGGGCGCG", "CACTAGCTAGCTACGCGCGAAAAACACTTTCACAG")]GRanges object with 4 ranges and 12 metadata columns:
seqnames ranges strand | qname flag mapq
<Rle> <IRanges> <Rle> | <character> <character> <character>
[1] allele1 0-24 + | myquery 0 0
[2] allele2 0-24 + | myquery 256 0
[3] allele2 27-48 + | myquery 0 0
[4] allele1 27-48 + | myquery 256 0
cigar rnext pnext tlen
<character> <character> <character> <character>
[1] 25M 0 -1 0
[2] 25M 0 -1 0
[3] 13S22M 0 -1 0
[4] 13S22M 0 -1 0
seq qual AS DD
<character> <character> <integer> <integer>
[1] CACTAGCTAGCTACGCGGGGGCGCG * 25 0
[2] CACTAGCTAGCTACGCGGGGGCGCG * 25 0
[3] CACTAGCTAGCTACGCGCGAAAAACACTTTCACAG * 22 0
[4] CACTAGCTAGCTACGCGCGAAAAACACTTTCACAG * 22 0
qwidth
<integer>
[1] 25
[2] 25
[3] 35
[4] 35
-------
seqinfo: 2 sequences from an unspecified genome; no seqlengths
## we can also use a pre existing fasta index
## this mini-genome was built using a chunk of hg19 chromosome 2
## and bwa index was run on the resulting .fasta
## (yielding files with .amb, .ann. .bwt, .pac, .sa extensions)
bwa = BWA(system.file('ext/mini.genome.fasta', package = 'RSeqLib'))
## inspect the bwa object
bwa
RSeqLib BWA object with params mc.cores = 1, hardclip = 0, keep_sec_with_frac_of_primary_score = 0.9, max_secondary = 10
## can retrieve the fasta path or character vector
## corresponding of the reference of the bwa object using genome(bwa)
genome(bwa)[1] "/gpfs/commons/groups/imielinski_lab/git/mskilab/rSeqLib/inst/ext/mini.genome.fasta"
## now let's query a character vector against this fasta reference
qstring = c("TGCTCTGGCACAAAGATAGGCATGCTCAGCCCACCTCTGCCAGCACTGGGGCTGAAGACTGGCCCACGTGTCCTCTCAATCCCCAGGACAACTTCACTATGGCTTTCATTAATAA",
"GCACAGATATCAACATAAGGACACAGAAAGCAGGAAAAGCAAGGAAATATGACTCCTTCAAAGGAACACAATAATTTTTCAGCATTAGATCTTAATCAGAAAGAACTTACTCCCAGATAAATAATTCAAAATAA")
## again we use the "[" operator to retrieve results
bwa[qstring]GRanges object with 2 ranges and 12 metadata columns:
seqnames ranges strand | qname flag mapq
<Rle> <IRanges> <Rle> | <character> <character> <character>
[1] 1 108294-108408 - | myquery 16 60
[2] 1 107951-108090 - | myquery 16 60
cigar rnext pnext tlen
<character> <character> <character> <character>
[1] 115M 0 -1 0
[2] 24M6D110M 0 -1 0
seq
<character>
[1] TTATTAATGAAAGCCATAGTGAAGTTGTCCTGGGGATTGAGAGGACACGTGGGCCAGTCTTCAGCCCCAGTGCTGGCAGAGGTGGGCTGAGCATGCCTATCTTTGTGCCAGAGCA
[2] TTATTTTGAATTATTTATCTGGGAGTAAGTTCTTTCTGATTAAGATCTAATGCTGAAAAATTATTGTGTTCCTTTGAAGGAGTCATATTTCCTTGCTTTTCCTGCTTTCTGTGTCCTTATGTTGATATCTGTGC
qual AS DD qwidth
<character> <integer> <integer> <integer>
[1] * 115 0 115
[2] * 122 0 134
-------
seqinfo: 1 sequence from an unspecified genome; no seqlengths
## character vector of "reads"
reads =
c('CTGCTAAGGAGATGGTGTGGCATCTGTGGAAGTCTGCTTGCACACAGCTGGAAGCCTTCAGGGGAAATAACAAACTGCTTCTCTTGCTGGGATTCCGACAT',
'CTAAGGAGATGGTGTGGCATCTGTGGAAGTCTGCTTGCACACAGCTGGAAGCCTTCAGGGGAAATAACAAACTGCTTCTCTTGCTGGGATTCCGACATGTC',
'TAAGGAGATGGTGTGGCATCTGTGGAAGTCTGCTTGCACACAGCTGGAAGCCTTCAGGGGAAATAACAAACTGCTTCTCTTGCTGGGATTCCGACATGTCC',
'GATGGTGTGGCATCTGTGGAAGTCTGCTTGCACACAGCTGGAAGCCTTCAGGGGAAATAACAAACTGCTTCTCTTGCTGGGATTCCGACATGTCCAAATAT',
'GGTGTGGCATCTGTGGAAGTCTGCTTGCACACAGCTGGAAGCCTTCAGGGGAAATAACAAACTGCTTCTCTTGCTGGGATTCCGACATGTCCAAATATGTC')
fermi = Fermi(reads, assemble = TRUE)
contigs(fermi)[1] "GACATATTTGGACATGTCGGAATCCCAGCAAGAGAAGCAGTTTGTTATTTCCCCTGAAGGCTTCCAGCTGTGTGCAAGCAGACTTCCACAGATGCCACACCATCTCCTTAGCAG"
## can also assemble a GRanges of reads (or any object
## coercible to a data.frame, or data.table)
## input just needs $seq, (optional) $qual, and (optional) $qname fields
## (so fermi takes base qualities into account)
reads = readRDS(system.file('ext/reads.rds', package = 'RSeqLib'))
names(as.data.table(reads)) [1] "seqnames" "start" "end" "width" "strand" "qname"
[7] "flag" "qwidth" "mapq" "cigar" "mrnm" "mpos"
[13] "isize" "seq" "qual" "MD" "MQ"
## make fermi object and assemble with error correction
fermi = Fermi(reads, assemble = TRUE)
fermiRSeqLib Fermi object with 6365 reads and 43 contigs:
## retrieve the contigs and align back to the reference via BWA object from mini.genome.fasta
gr = bwa[contigs(fermi)]
## the first contig alignment comprises 732 bases matching the reference
gr[1]
GRanges object with 1 range and 12 metadata columns:
seqnames ranges strand | qname flag mapq
<Rle> <IRanges> <Rle> | <character> <character> <character>
[1] chr2 107677-108408 - | myquery 16 60
cigar rnext pnext tlen
<character> <character> <character> <character>
[1] 732M 0 -1 0
seq
<character>
[1] TTTTTTTTTTAATTGTCTGATTGGAGTATTTAAAAAGATCTGTCTTCAAGTTCTGAGATTCTTTATTCTACCTGATCTAATCTATTGTTGATGCTTTCAAATGTGTTTTTTTTATTTCCTTCAATGAATTCTTCAGTTCCAGAATTTCTATTTGGTTCATTTAAAAATGTCTATTTCTGTGTAAATTTCTCATTTATATTCTGAATTGATTTTCTAATTTCTTTCTATTGTTTTTCAGAATTTTCTTGTATTTTACTGAGTTTCTTTAAAATCATTATTTTGAATTATTTATCTGGGATGTTATGTAAGTTCTTTCTGATTAAGATCTAATGCTGAAAAATTATTGTGTTCCTTTGAAGGAGTCATATTTCCTTGCTTTTCCTGCTTTCTGTGTCCTTATGTTGATATCTGTGCATCTGGCATAATAGTCACCTTCTATTTTTGAATTTGCTTTTATAGGGGAGGACTTTTTCCTGAAGATGTATCTCTGGTATCAGTTGGGTAGAGCACTTTGGCTTTGATTCTGGGTGCATCCAGTAGTGTAGTCTCCATATGATTTCTTTGGCTGTAAACAGTGTTAGTAGCATCTGTGATTTCCTTCATGGATTAGAGTGTGGTTATTAATGAAAGCCATAGTGAAGTTGTCCTGGGGATTGAGAGGACACGTGGGCCAGTCTTCAGCCCCAGTGCTGGCAGAGGTGGGCTGAGCATGCCTATCTTTGTGCCAGAGCA
qual AS DD qwidth
<character> <integer> <integer> <integer>
[1] * 732 0 732
Marcin Imielinski - Assistant Professor, Weill Cornell Medicine Core Member, New York Genome Center
Jeremiah Wala - MD PhD Student, Harvard / MIT / HST
Khagay Nagdimov - Intern, Imielinski Lab, New York Genome Center