Bio.Data.Bed.Utils

Documentation

fetchSeq :: (BioSeq DNA a, MonadIO m) => Genome -> ConduitT BED (Either String (DNA a)) m ()

retreive sequences

fetchSeq' :: (BioSeq DNA a, MonadIO m) => Genome -> [BED] -> m [Either String (DNA a)]

motifScan :: (BEDLike b, MonadIO m) => Genome -> [Motif] -> Bkgd -> Double -> ConduitT b BED m ()

Identify motif binding sites

getMotifScore :: MonadIO m => Genome -> [Motif] -> Bkgd -> ConduitT BED BED m ()

Retrieve motif matching scores

getMotifPValue

Arguments

:: Monad m

=> Maybe Double

whether to truncate the motif score CDF. Doing this will significantly reduce memory usage without sacrifice accuracy.

-> [Motif]

-> Bkgd

-> ConduitT BED BED m ()

monoColonalize :: Monad m => ConduitT BED BED m ()

process a sorted BED stream, keep only mono-colonal tags

newtype BaseMap

Constructors

BaseMap (HashMap ByteString BitVector)

baseMap

Arguments

:: PrimMonad m

=> [(ByteString, Int)]

chromosomes and their sizes

-> ConduitT BED o m BaseMap

Count the tags (starting positions) at each position in the genome.

queryBaseMap :: BEDLike b => b -> BaseMap -> Maybe [Bool]

rpkmBed :: (PrimMonad m, BEDLike b, Vector v Double) => [b] -> ConduitT BED o m (v Double)

calculate RPKM on a set of unique regions. Regions (in bed format) would be kept in memory but not tag file. RPKM: Readcounts per kilobase per million reads. Only counts the starts of tags

rpkmSortedBed :: (PrimMonad m, BEDLike b, Vector v Double) => Sorted (Vector b) -> ConduitT BED o m (v Double)

calculate RPKM on a set of regions. Regions must be sorted. The Sorted data type is used to remind users to sort their data.

countTagsBinBed

Arguments

:: (Integral a, PrimMonad m, Vector v a, BEDLike b)

=> Int

bin size

-> [b]

regions

-> ConduitT BED o m ([v a], Int)

divide each region into consecutive bins, and count tags for each bin and return the number of all tags. Note: a tag is considered to be overlapped with a region only if the starting position of the tag is in the region. For the common sense overlapping, use countTagsBinBed'.

countTagsBinBed'

Arguments

:: (Integral a, PrimMonad m, Vector v a, BEDLike b1, BEDLike b2)

=> Int

bin size

-> [b1]

regions

-> ConduitT b2 o m ([v a], Int)

Same as countTagsBinBed, except that tags are treated as complete intervals instead of single points.

tagCountDistr :: PrimMonad m => Vector v Int => ConduitT BED o m (v Int)

peakCluster

Arguments

:: (BEDLike b, Monad m)

=> [b]

peaks

-> Int

radius

-> Int

cutoff

-> ConduitT o BED m ()

cluster peaks