diff --git a/Biobase/Primary.hs b/Biobase/Primary.hs --- a/Biobase/Primary.hs +++ b/Biobase/Primary.hs @@ -32,16 +32,13 @@ import qualified Data.Vector.Generic.Mutable as VGM import qualified Data.Vector.Unboxed as VU -import Data.ExtShape -import Data.PrimitiveArray -import Data.PrimitiveArray.Zero.Unboxed -import "PrimitiveArray" Data.Array.Repa.Index -import "PrimitiveArray" Data.Array.Repa.Shape +import Data.Array.Repa.ExtShape +import Data.Array.Repa.Index +import Data.Array.Repa.Shape import Biobase.Primary.Bounds - -- * Convert different types of sequence representations to the internal -- "Primary Structure" representation @@ -51,24 +48,11 @@ class MkPrimary a where mkPrimary :: a -> Primary -type Primary = Arr0 DIM1 Nuc +type Primary = VU.Vector Nuc -instance Eq Primary where - xs == ys - | bx==by = sliceEq xs zeroDim ys zeroDim bx - | otherwise = False - where (_,bx) = bounds xs - (_,by) = bounds ys -instance Ord Primary where - xs <= ys - | bx==by = toList xs <= toList ys - | otherwise = bx<=by - where (_,Z:.bx) = bounds xs - (_,Z:.by) = bounds ys - -- * Efficient nucleotide encoding -- A 'Nuc'leotide is simply an Int wrapped up. 'nIMI' provides for @@ -78,13 +62,13 @@ newtype Nuc = Nuc {unNuc :: Int} deriving (Eq,Ord,Ix) -(nN : nA : nC : nG : nT : nIMI : nUndefined : _) = map Nuc [0 .. ] -nU = nT +(nN : nA : nC : nG : nT : nU : nIMI : nUndefined : _) = map Nuc [0 .. ] -acgt = [nA..nT] -acgu = acgt -nacgt = [nN..nT] -nacgu = nacgt +acgt = [nA,nC,nG,nT] +acgu = [nA,nC,nG,nU] +cgau = [nC,nG,nA,nU] +nacgt = nN:acgt +nacgu = nN:acgu -- | Translate between 'Char's and 'Nuc's. @@ -180,6 +164,7 @@ subDim (sh1:.Nuc n1) (sh2:.Nuc n2) = subDim sh1 sh2 :. Nuc (n1-n2) rangeList (sh1:.Nuc n1) (sh2:.Nuc n2) = [ sh:.Nuc n | sh <- rangeList sh1 sh2, n <- [n1 .. (n1+n2)]] +{- -- | The bounded instance from GHC proper. Captures all defined symbols. instance Bounded Nuc where @@ -194,6 +179,9 @@ minExtended = nN maxExtended = nT + +-} + -- | Enum instance Enum Nuc where @@ -203,7 +191,7 @@ -- ** Instances for 'MkPrimary' instance MkPrimary String where - mkPrimary xs = fromList (Z:.0) (Z:.length xs -1) $ map mkNuc xs + mkPrimary = VU.fromList . map mkNuc instance MkPrimary BS.ByteString where mkPrimary = mkPrimary . BS.unpack @@ -215,5 +203,5 @@ mkPrimary = mkPrimary . T.unpack instance MkPrimary [Nuc] where - mkPrimary xs = fromList (Z:.0) (Z:.length xs -1) xs + mkPrimary = VU.fromList diff --git a/Biobase/Primary/Hashed.hs b/Biobase/Primary/Hashed.hs --- a/Biobase/Primary/Hashed.hs +++ b/Biobase/Primary/Hashed.hs @@ -14,8 +14,6 @@ import qualified Data.Vector.Generic.Mutable as VGM import qualified Data.Vector.Unboxed as VU -import Data.PrimitiveArray - import Biobase.Primary @@ -42,9 +40,9 @@ -- for 'ps'. We could in O(1) create a vector from a Primary ... mkHashedPrimary :: (Nuc,Nuc) -> Primary -> HashedPrimary -mkHashedPrimary (l,u) ps' = assert (VU.all (\p -> l<=p && p<=u) ps) $ HashedPrimary idx where +mkHashedPrimary (l,u) ps = assert (VU.all (\p -> l<=p && p<=u) ps) $ HashedPrimary idx where idx = VU.sum $ VU.zipWith f ps (VU.enumFromStepN (VU.length ps -1) (-1) (VU.length ps)) f p c = (unNuc p - unNuc l) * (cnst^c) cnst = unNuc u - unNuc l + 1 - ps = VU.fromList $ toList ps' {-# INLINE mkHashedPrimary #-} + diff --git a/Biobase/Secondary.hs b/Biobase/Secondary.hs --- a/Biobase/Secondary.hs +++ b/Biobase/Secondary.hs @@ -17,8 +17,8 @@ module Biobase.Secondary where -import "PrimitiveArray" Data.Array.Repa.Index -import "PrimitiveArray" Data.Array.Repa.Shape +import Data.Array.Repa.Index +import Data.Array.Repa.Shape import Data.Char (toLower, toUpper) import Data.Ix (Ix(..)) import Data.List as L @@ -346,3 +346,4 @@ {-# INLINE updR #-} {-# INLINE updP #-} {-# INLINE updT #-} + diff --git a/Biobase/Secondary/Constraint.hs b/Biobase/Secondary/Constraint.hs --- a/Biobase/Secondary/Constraint.hs +++ b/Biobase/Secondary/Constraint.hs @@ -7,16 +7,17 @@ module Biobase.Secondary.Constraint where +import Data.Array.Repa.Index +import Data.Array.Repa.Shape import Data.Char (toLower) import Data.Primitive.Types import qualified Data.Vector.Generic as VG import qualified Data.Vector.Generic.Mutable as VGM import qualified Data.Vector.Unboxed as VU +import Prelude as P import Data.PrimitiveArray -import Data.PrimitiveArray.Zero.Unboxed -import "PrimitiveArray" Data.Array.Repa.Index -import "PrimitiveArray" Data.Array.Repa.Shape +import Data.PrimitiveArray.Zero import Biobase.Secondary.Diagrams @@ -51,7 +52,7 @@ -- TODO and again, we should parametrize over "Energy", "Score", etc (that is, -- Prim a) -bonusTable :: Double -> Double -> Constraint -> Arr0 DIM2 Double +bonusTable :: Double -> Double -> Constraint -> Unboxed DIM2 Double bonusTable bonus malus (Constraint constraint) = arr where arr = fromAssocs zeroDim (Z:.n:.n) 0 $ bonusBr ++ bonusAn ++ bonusBa ++ malusBr ++ malusAn ++ malusX n = VU.length constraint -1 @@ -109,8 +110,9 @@ -- * Instances instance MkConstraint String where - mkConstraint xs = mkConstraint . VU.fromList . map toLower $ xs + mkConstraint xs = mkConstraint . VU.fromList . P.map toLower $ xs instance MkConstraint (VU.Vector Char) where mkConstraint cs = Constraint $ VU.zip cs ks where (D1S ks) = mkD1S cs + diff --git a/Biobase/Secondary/Diagrams.hs b/Biobase/Secondary/Diagrams.hs --- a/Biobase/Secondary/Diagrams.hs +++ b/Biobase/Secondary/Diagrams.hs @@ -4,17 +4,12 @@ -- | Types for RNA secondary structure. Types vary from the simplest array -- (D1Secondary) to rather complex ones. --- --- TODO The complex ones are still coming in from other libraries. --- --- TODO can we use Char8 instead of Char? --- --- TODO prepare for extended RNA secondary structures! {-# LANGUAGE RecordWildCards #-} module Biobase.Secondary.Diagrams where + import qualified Data.Vector.Unboxed as VU import Data.List (sort,groupBy,sortBy) import Data.Tuple.Select (sel1,sel2) @@ -90,6 +85,7 @@ d2Grouping ((i,j),_) ((k,l),_) = i<=k && j>=l +{- test :: (Int,[ExtPairIdx]) test = (20,test') @@ -102,7 +98,7 @@ , ((2,18),(trans,sugar,sugar)) , ((15,18),(cis,sugar,sugar)) ] - +-} -- * Instances for D1S @@ -200,3 +196,4 @@ g k (s:st) (x:xs) | r==x = (s,k) : g (k+1) st xs g a b c = error $ show (a,b,c) + diff --git a/Biobase/Secondary/PseudoKnots.hs b/Biobase/Secondary/PseudoKnots.hs --- a/Biobase/Secondary/PseudoKnots.hs +++ b/Biobase/Secondary/PseudoKnots.hs @@ -59,3 +59,4 @@ instance RemovePseudoKnots [ExtPairIdx] where removeByCounting = VU.toList . removeByCounting . VU.fromList + diff --git a/Biobase/Secondary/Vienna.hs b/Biobase/Secondary/Vienna.hs --- a/Biobase/Secondary/Vienna.hs +++ b/Biobase/Secondary/Vienna.hs @@ -10,6 +10,8 @@ module Biobase.Secondary.Vienna where +import Data.Array.Repa.Index +import Data.Array.Repa.Shape import Data.Ix import Data.Primitive.Types import Data.Tuple (swap) @@ -17,12 +19,11 @@ import qualified Data.Vector.Generic as VG import qualified Data.Vector.Generic.Mutable as VGM import qualified Data.Vector.Unboxed as VU +import Prelude as P -import Data.ExtShape +import Data.Array.Repa.ExtShape import Data.PrimitiveArray as PA -import Data.PrimitiveArray.Zero.Unboxed as PA -import "PrimitiveArray" Data.Array.Repa.Index -import "PrimitiveArray" Data.Array.Repa.Shape +import Data.PrimitiveArray.Zero as PA import Biobase.Primary import Biobase.Primary.Bounds @@ -70,15 +71,14 @@ subDim (sh1:.ViennaPair n1) (sh2:.ViennaPair n2) = subDim sh1 sh2 :. (ViennaPair $ n1-n2) rangeList (sh1:.ViennaPair n1) (sh2:.ViennaPair n2) = [sh:.ViennaPair n | sh <- rangeList sh1 sh2, n <- [n1 .. (n1+n2)]] -(vpNP:vpCG:vpGC:vpGU:vpUG:vpAU:vpUA:vpNS:vpUndefined:_) = map ViennaPair [0..] +(vpNP:vpCG:vpGC:vpGU:vpUG:vpAU:vpUA:vpNS:vpUndefined:_) = P.map ViennaPair [0..] class MkViennaPair a where mkViennaPair :: a -> ViennaPair fromViennaPair :: ViennaPair -> a instance MkViennaPair (Nuc,Nuc) where - mkViennaPair (b1,b2) = viennaPairTable `PA.index` (Z:.b1:.b2) - {- + mkViennaPair (b1,b2) -- = viennaPairTable `PA.index` (Z:.b1:.b2) | b1==nC&&b2==nG = vpCG | b1==nG&&b2==nC = vpGC | b1==nG&&b2==nU = vpGU @@ -86,7 +86,6 @@ | b1==nA&&b2==nU = vpAU | b1==nU&&b2==nA = vpUA | otherwise = vpNS - -} {-# INLINE mkViennaPair #-} fromViennaPair p | p==vpCG = (nC,nG) @@ -98,7 +97,18 @@ | otherwise = error "non-standard pairs can't be backcasted" {-# INLINE fromViennaPair #-} -viennaPairTable :: Arr0 (Z:.Nuc:.Nuc) ViennaPair +isViennaPair :: Nuc -> Nuc -> Bool +isViennaPair a b = f a b where + f l r = l==nC && r==nG + || l==nG && r==nC + || l==nA && r==nU + || l==nU && r==nA + || l==nG && r==nU + || l==nU && r==nG + {-# INLINE f #-} +{-# INLINE isViennaPair #-} + +viennaPairTable :: Unboxed (Z:.Nuc:.Nuc) ViennaPair viennaPairTable = fromAssocs (Z:.nN:.nN) (Z:.nU:.nU) vpNS [ (Z:.nC:.nG , vpCG) , (Z:.nG:.nC , vpGC) @@ -144,7 +154,7 @@ | x ==' ' = readsPrec p (y:xs) | Just n <- (x:y:[]) `lookup` s2p = [(n,xs)] | otherwise = [] - where s2p = (map swap pairToString) + where s2p = (P.map swap pairToString) @@ -168,3 +178,4 @@ cguaP = [vpCG..vpUA] cgnsP = [vpCG..vpNS] pairToString = [(vpCG,"CG"),(vpGC,"GC"),(vpUA,"UA"),(vpAU,"AU"),(vpGU,"GU"),(vpUG,"UG"),(vpNS,"NS"),(vpNP,"NP")] + diff --git a/BiobaseXNA.cabal b/BiobaseXNA.cabal --- a/BiobaseXNA.cabal +++ b/BiobaseXNA.cabal @@ -1,9 +1,9 @@ name: BiobaseXNA -version: 0.6.3.1 +version: 0.7.0.0 author: Christian Hoener zu Siederdissen maintainer: choener@tbi.univie.ac.at homepage: http://www.tbi.univie.ac.at/~choener/ -copyright: Christian Hoener zu Siederdissen, 2011-2012 +copyright: Christian Hoener zu Siederdissen, 2011-2013 category: Bioinformatics synopsis: Efficient RNA/DNA representations license: GPL-3 @@ -13,11 +13,14 @@ cabal-version: >= 1.6.0 description: This is a base library for bioinformatics with emphasis on RNA - and DNA primary structure and related tools. Provided are - efficient encodings for short sequences, as required by RNA - folding tools. Extended RNA secondary structures can be - represented as well. + and DNA primary structure. . + Provided are efficient encodings for short sequences, as + required by RNA folding tools. Extended RNA secondary + structures can be represented as well. + . + . + . Contains data from: . Frequency and isostericity of RNA base pairs @@ -30,8 +33,12 @@ . . . - New in 0.6.2.0 + new in 0.7 . + * updated to PrimitiveArray >= 0.5 + . + new in 0.6.2.0 + . * Updated to PrimitiveArray >= 0.2.0.0 extra-source-files: @@ -41,15 +48,16 @@ library build-depends: base >3 && <5, - bytestring >= 0.9 , + bytestring >= 0.10 , containers >= 0.4 , csv == 0.1.2 , - file-embed >= 0.0.4 , - primitive >= 0.4 , + file-embed >= 0.0.4.7 , + primitive >= 0.5 , + repa >= 3.2 , text >= 0.11 , tuple >= 0.2 , vector >= 0.10 , - PrimitiveArray >= 0.4.0.1 + PrimitiveArray >= 0.5 exposed-modules: Biobase.Primary