diff --git a/Biobase/Infernal/CM.hs b/Biobase/Infernal/CM.hs new file mode 100644 --- /dev/null +++ b/Biobase/Infernal/CM.hs @@ -0,0 +1,53 @@ + +-- | Infernal CMs. + +module Biobase.Infernal.CM where + +import Data.Vector as V +import Data.Vector.Unboxed as VU +import Data.ByteString as BS + +import Data.PrimitiveArray +import Data.PrimitiveArray.Ix + + + +-- | A datatype representing Infernal covariance models. This is a new +-- representation that is incompatible with the one once found in "Biobase". +-- The most important difference is that lookups are mapped onto efficient data +-- structures, currently "PrimitiveArray". +-- +-- [1] Each "State" of a covariance model has up to 6 transition scores, hence +-- we need s*6 cells for transitions. +-- +-- [2] Each "State" of a covariance has up to 16 emission scores, so we have +-- s*16 cells for emissions, with unused cells set to a really high score. +-- +-- On top of these basic structures, we then place additional high-level +-- constructs. +-- +-- [3] 'paths' are allowed transitions. This can safe a check, if the +-- transition is encoded with a forbidden score. +-- +-- [4] 'localBegin' and 'localEnd' are local entry and exit strategies. A +-- 'localBegin' is a transition score to certain states, all such transitions +-- are in 'begins'. A 'localEnd' is a transition score to a local end state. +-- +-- TODO as with other projects, we should not use Double's but "Score" and +-- "Probability" newtypes. + +data CM = CM + { name :: ByteString + , accession :: Int + , ga :: Double + , tc :: Double + , nc :: Double + , transition :: PrimArray (Int,Int) Double + , emission :: PrimArray (Int,Int) Double + , paths :: V.Vector (VU.Vector Double) + , localBegin :: VU.Vector Double + , begins :: VU.Vector Int + , localEnd :: VU.Vector (Double) + , nodes :: V.Vector (VU.Vector Int) + } + deriving (Show) diff --git a/Biobase/Infernal/CM/Export.hs b/Biobase/Infernal/CM/Export.hs new file mode 100644 --- /dev/null +++ b/Biobase/Infernal/CM/Export.hs @@ -0,0 +1,12 @@ + +-- | Transforms the internal representation of a CM back into a version that +-- can be used by Infernal. +-- +-- Note that models are transformed into 'ByteString' as-is, the exporter does +-- not make sure that probabilities add to one, that we write out probabilities +-- instead of scores, and so on. +-- +-- TODO some of the notes above will become less problematic once we use +-- newtypes, as a probability-CM will not be accepted by the exporter by then. + +module Biobase.Infernal.CM.Export where diff --git a/Biobase/Infernal/CM/Import.hs b/Biobase/Infernal/CM/Import.hs new file mode 100644 --- /dev/null +++ b/Biobase/Infernal/CM/Import.hs @@ -0,0 +1,34 @@ + +-- | Iteratee-based parsing of Infernal covariance models. + +module Biobase.Infernal.CM.Import where + +import Data.Iteratee as I +import Data.Iteratee.Char as I +import Data.Iteratee.IO as I +import Data.Iteratee.Iteratee as I +import Data.Iteratee.ListLike as I +import Data.ByteString as BS + +import Biobase.Infernal.CM + + + +-- * iteratee stuff + +-- | + +eneeCM :: (Monad m) => Enumeratee ByteString [CM] m a +eneeCM = enumLinesBS ><> convStream f where + f = do + th <- tryHead + return undefined + + +-- * convenience functions + +-- | Read covariance models from file. This parser reads one or more CMs from +-- file. + +fromFile :: FilePath -> [CM] +fromFile = undefined diff --git a/Biobase/Infernal/VerboseHit.hs b/Biobase/Infernal/VerboseHit.hs --- a/Biobase/Infernal/VerboseHit.hs +++ b/Biobase/Infernal/VerboseHit.hs @@ -1,4 +1,5 @@ {-# LANGUAGE RecordWildCards #-} +{-# OPTIONS_GHC -funbox-strict-fields #-} -- | Provides a datatype for cmsearch verbose output. The Import/Export system -- now allows for primitive annotations using "##" as the first two characters. @@ -14,20 +15,20 @@ -- | Captures a complete alignment data VerboseHit = VerboseHit - { vhTarget :: (Int,Int) -- ^ part of target sequence (start counting at 1) - , vhQuery :: (Int,Int) -- ^ which part of the CM/stk do we align to - , vhCM :: ByteString -- ^ the CM for this alignment - , vhStrand :: Strand -- ^ should be either '+' or '-' - , vhScore :: Double -- ^ bit score - , vhEvalue :: Double -- ^ number of hits we expect to find with 'score' or higher for 'targetSequence' length - , vhPvalue :: Double -- ^ ? - , vhGC :: Int -- ^ ? - , vhScaffold :: ByteString -- ^ scaffold, chromosome, ... (the name of the sequence, not the sequence data!) - , vhWuss :: ByteString -- ^ fancy secondary structure annotation using wuss notation - , vhConsensus :: ByteString -- ^ query consensus (upper: highly, lower: weak/no) - , vhScoring :: ByteString -- ^ represents where positive and negative scores come from - , vhSequence :: ByteString -- ^ the target sequence which aligns to the model - , vhAnnotation :: [ByteString] -- ^ any annotations that could be associated (# lines) + { vhTarget :: !(Int,Int) -- ^ part of target sequence (start counting at 1) + , vhQuery :: !(Int,Int) -- ^ which part of the CM/stk do we align to + , vhCM :: !ByteString -- ^ the CM for this alignment + , vhStrand :: !Strand -- ^ should be either '+' or '-' + , vhScore :: !Double -- ^ bit score + , vhEvalue :: !Double -- ^ number of hits we expect to find with 'score' or higher for 'targetSequence' length + , vhPvalue :: !Double -- ^ ? + , vhGC :: !Int -- ^ ? + , vhScaffold :: !ByteString -- ^ scaffold, chromosome, ... (the name of the sequence, not the sequence data!) + , vhWuss :: !ByteString -- ^ fancy secondary structure annotation using wuss notation + , vhConsensus :: !ByteString -- ^ query consensus (upper: highly, lower: weak/no) + , vhScoring :: !ByteString -- ^ represents where positive and negative scores come from + , vhSequence :: !ByteString -- ^ the target sequence which aligns to the model + , vhAnnotation :: ![ByteString] -- ^ any annotations that could be associated (# lines) } deriving (Show,Read) type Strand = Char diff --git a/Biobase/Infernal/VerboseHit/Export.hs b/Biobase/Infernal/VerboseHit/Export.hs --- a/Biobase/Infernal/VerboseHit/Export.hs +++ b/Biobase/Infernal/VerboseHit/Export.hs @@ -1,3 +1,4 @@ +{-# LANGUAGE NoMonomorphismRestriction #-} {-# LANGUAGE RankNTypes #-} {-# LANGUAGE RecordWildCards #-} {-# LANGUAGE OverloadedStrings #-} @@ -39,8 +40,9 @@ f acc = do h <- I.head let na = newAcc acc h + p <- I.peek return ( fst na - , return . BS.unlines $ snd na ++ P.map (append "## ") (vhAnnotation h) ++ [showVerboseHit h] + , return . BS.unlines $ snd na ++ P.map (append "##") (vhAnnotation h) ++ [showVerboseHit h] ++ maybe ["//"] (const []) p ) -- | Given the current state "a" and verbose hit "h", determine if any state @@ -48,10 +50,13 @@ newAcc a@(AliGo{..}) h@VerboseHit{..} | otherwise = ( AliGo vhCM vhScaffold vhStrand [], ls ) - where ls = [ "//" | aliCM /= BS.empty && aliCM /= vhCM ] ++ - [ "CM: " `BS.append` vhCM | aliCM /= vhCM ] ++ - [ ">" `BS.append` vhScaffold `BS.append` "\n" | aliScaffold /= vhScaffold ] ++ - [ str `BS.append` " strand results:\n" | aliStrand /= vhStrand ] + where ls = [ "//" | aliCM /= BS.empty && bCM ] ++ + [ "CM: " `BS.append` vhCM | bCM ] ++ + [ ">" `BS.append` vhScaffold `BS.append` "\n" | bCM || bSc] ++ + [ str `BS.append` " strand results:\n" | bCM || bSc || bSt ] + bCM = aliCM /= vhCM + bSc = aliScaffold /= vhScaffold + bSt = aliStrand /= vhStrand str | vhStrand == '+' = "Plus" | vhStrand == '-' = "Minus" @@ -83,14 +88,19 @@ {- -import Biobase.Infernal.VerboseHit.Import +--import Biobase.Infernal.VerboseHit.Import test = do xs <- fromFile "/home/choener/tmp/infernal-1.0.2/tutorial/tmp.res" i <- enumList [xs] $ joinI $ eneeByteString stream2stream ys <- run i + {- BS.putStrLn ys print $ BS.length ys print $ P.length $ BS.lines ys + -} + BS.putStrLn $ BS.take 1000 ys return () -} + + diff --git a/Biobase/Infernal/VerboseHit/Import.hs b/Biobase/Infernal/VerboseHit/Import.hs --- a/Biobase/Infernal/VerboseHit/Import.hs +++ b/Biobase/Infernal/VerboseHit/Import.hs @@ -41,12 +41,12 @@ case h' of Nothing -> return (acc, []) (Just h) - | "##" `isPrefixOf` h -> return (acc{aliAnnotation = aliAnnotation acc ++ [BS.drop 2 h]},[]) - | "CM: " `isInfixOf` h -> return (acc{aliCM = BS.copy $ BS.drop 4 h, aliAnnotation = []}, []) - | ">" `isInfixOf` h -> return (acc{aliScaffold = BS.copy $ BS.drop 1 h, aliAnnotation = []}, []) - | " Plus" `isInfixOf` h -> return (acc{aliStrand = '+', aliAnnotation = []}, []) - | " Minus" `isInfixOf` h -> return (acc{aliStrand = '-', aliAnnotation = []}, []) - | " Query" `isInfixOf` h -> do + | "##" `isPrefixOf` h -> return (acc{aliAnnotation = aliAnnotation acc ++ [BS.drop 2 h]},[]) + | "CM: " `isPrefixOf` h -> return (acc{aliCM = BS.copy $ BS.drop 4 h, aliAnnotation = []}, []) + | ">" `isPrefixOf` h -> return (acc{aliScaffold = BS.copy $ BS.drop 1 h, aliAnnotation = []}, []) + | "Plus strand results" `isInfixOf` h -> return (acc{aliStrand = '+', aliAnnotation = []}, []) + | "Minus strand results" `isInfixOf` h -> return (acc{aliStrand = '-', aliAnnotation = []}, []) + | " Query" `isInfixOf` h -> do x <- qs h (aliCM acc) (aliScaffold acc) (aliStrand acc) (aliAnnotation acc) return (acc{aliAnnotation = []},x) | otherwise -> return (acc,[]) @@ -130,9 +130,10 @@ -- How to use this enumeratee. +{- test = do i <- enumFile 8192 "test.vh" $ joinI $ eneeVerboseHit stream2list xs <- run i P.mapM_ (\x -> print x >> P.putStrLn "\n\n\n") xs print $ P.length xs - +-} diff --git a/BiobaseInfernal.cabal b/BiobaseInfernal.cabal --- a/BiobaseInfernal.cabal +++ b/BiobaseInfernal.cabal @@ -1,5 +1,5 @@ name: BiobaseInfernal -version: 0.5.4.0 +version: 0.5.4.1 author: Christian Hoener zu Siederdissen maintainer: choener@tbi.univie.ac.at homepage: http://www.tbi.univie.ac.at/~choener/ @@ -22,13 +22,11 @@ . . . - With the BioHaskell library split, this package is officially - back! And the package is not feature-complete yet, take the - above as a will-include-soon list. + Changes in 0.5.4.1 . - The taxonomy importer makes use of Iteratee.zip, hence the - switch from Enumerator. (See the Biohaskell wiki pages for - discussion on Iteratee/Enumerator). + * fix-up for VH export + . + * half-baked CM type (not useful yet) extra-source-files: @@ -42,12 +40,17 @@ either-unwrap, iteratee, transformers, - tuple + tuple, + vector, + PrimitiveArray exposed-modules: Biobase.Infernal Biobase.Infernal.Clan Biobase.Infernal.Clan.Import + Biobase.Infernal.CM + Biobase.Infernal.CM.Export + Biobase.Infernal.CM.Import Biobase.Infernal.TabularHit Biobase.Infernal.TabularHit.Import Biobase.Infernal.Taxonomy