From e6a83e28ef3e611cdba53098dcc17034607aa78f Mon Sep 17 00:00:00 2001
From: Jonathan Ott <jonathan.ott1994@gmail.com>
Date: Thu, 11 Jan 2024 16:21:01 +0100
Subject: [PATCH] add cwl compatible version and workflow
---
runs/TurboIDSampleArc.yml | 13 +
workflows/EvalTurboID/EvalTurboID.fsx | 696 ++++++++++++++++++++-
workflows/EvalTurboID/TurboIDSampleArc.cwl | 48 ++
3 files changed, 731 insertions(+), 26 deletions(-)
create mode 100644 runs/TurboIDSampleArc.yml
create mode 100644 workflows/EvalTurboID/TurboIDSampleArc.cwl
diff --git a/runs/TurboIDSampleArc.yml b/runs/TurboIDSampleArc.yml
new file mode 100644
index 0000000..87e4dae
--- /dev/null
+++ b/runs/TurboIDSampleArc.yml
@@ -0,0 +1,13 @@
+arcDirectory:
+ class: Directory
+ path: ../
+mappingFile:
+ class: File
+ path: ../workflows/EvalTurboID/scripts/chlamy_jgi55.txt
+scriptFile:
+ class: File
+ path: ../workflows/EvalTurboID/EvalTurboID.fsx
+input:
+ class: File
+ path: ../assays/MSEval/dataset/TEF_PSBQ_LON_MS264DDAfs_combined_protein.tsv
+runID: test
diff --git a/workflows/EvalTurboID/EvalTurboID.fsx b/workflows/EvalTurboID/EvalTurboID.fsx
index 452d083..09344df 100644
--- a/workflows/EvalTurboID/EvalTurboID.fsx
+++ b/workflows/EvalTurboID/EvalTurboID.fsx
@@ -19,18 +19,659 @@ open ISADotNet
open ISADotNet
open ISADotNet.QueryModel
-#load "scripts/annotate.fsx"
+module Annotate =
+
+ open System
+ open Argu
+ open Deedle
+
+ printfn "=====\nCheck if dotnet sdk reference can be rechanged to latest: https://stackoverflow.com/a/65934809 \n====="
+
+ /// \t or "\t" doesnt work as argument
+ let getSeparator str =
+ match str with
+ | "tab" -> "\t"
+ | "tabulator" -> "\t"
+ | _ -> str
+
+ /// Snapshots of MapMan and Gene Ontology were generated with FATool and genome releases 5.5 (Chlamy) and Araport11 (Arabidopsis)
+ module Mapping =
+
+ /// reads FA tool snapshot for araport11 or chlamyJGI_v5.5 or other annotation files
+ let readMapping (mappingFilePath:string) (columnSeparator:string) (identifierColHeader:string) :Frame<string,string> =
+ Frame.ReadCsv(mappingFilePath,hasHeaders=true,separators = columnSeparator)
+ |> Frame.indexRows identifierColHeader
+
+ /// truncateID processes identifier (e.g. Cre10.g123456.t2.1 -> Cre10.g123456)
+ let getAnnotationsFromIdentifier (frame:Frame<string,string>) (annotationHeader:string[]) (multipleIdentifierSeparator:string) (multipleAnnotationSeparator:string) (truncateID:string->string) (identifier:string) =
+
+ /// identifier that should be mapped
+ let identifier = identifier.Split([|multipleIdentifierSeparator|],StringSplitOptions.None)
+
+ /// Mappings from all identifer to annotations, that are present in annotation frame
+ let mappings :Series<string,string> []=
+ identifier
+ |> Array.choose (fun ident ->
+ let truncId = truncateID ident
+ let k = frame.TryGetRow truncId
+ if k.HasValue then
+ Some k.Value
+ else
+ printfn "Warning: The following id could not be found within mapping file: %s" truncId
+ None
+ )
+
+ /// all annotations that should be used
+ annotationHeader //["GO","Synonym"]
+ |> Array.map (fun annotationType ->
+ mappings //[[GO => GO:006; Synonym => Q0WV96];[GO => GO:001; Synonym => Q01337]]
+ |> Array.map (fun mapping ->
+ let annotation = mapping.[annotationType]
+ annotation.Split ';'
+ |> String.concat multipleAnnotationSeparator
+ )
+ )
+
+ let getAnnotationRow (frame:Frame<string,string>) (annotationHeader:string[]) multipleIdentifierSeparator multipleAnnotationSeparator truncateID (identifier:string) =
+ getAnnotationsFromIdentifier frame annotationHeader multipleIdentifierSeparator multipleAnnotationSeparator truncateID identifier
+ |> Array.map (String.concat multipleIdentifierSeparator)
+
+
+ /// User data is read, annotated and written to a new file
+ module Data =
+
+ let getDataFrame (columnSeparator:string) inputPath =
+ System.IO.File.ReadAllLines(inputPath)
+ |> Array.map (fun x ->
+ x.Split([|columnSeparator|],System.StringSplitOptions.None)
+ )
+
+ /// index of column that contains the identifier to annotate
+ let getColIndex (dataFrame:string[][]) columnHeader=
+ Array.tryFindIndex (fun x -> x = columnHeader) dataFrame.[0]
+ |> fun o ->
+ match o with
+ | Some i -> i
+ | _ -> failwithf "ColumnHeader %s not found." columnHeader
+
+ /// based on given mapping arguments the file is extended with given mapping columns
+ let getHeader (dataFrame:string[][]) (columnSeparator:string) (annotationHeader:string[]) =
+ Array.append dataFrame.[0] annotationHeader
+ |> String.concat columnSeparator
+
+ /// every row of the file is processed and converted to a new string with additional information attached at the end of the line
+ let getAnnotatedLines annotationFrame inputPath (columnSeparator:string) columnHeader (annotations:string[]) (multipleIdentifierSeparator:string) truncateID (multipleAnnotationSeparator:string) =
+ let dataFrame = getDataFrame columnSeparator inputPath
+ let colIndex = getColIndex dataFrame columnHeader
+ let header = getHeader dataFrame columnSeparator annotations
+ //let rowCount = dataFrame.Length
+ dataFrame
+ |> Array.tail
+ |> Array.mapi (fun i x ->
+ let identifier = x.[colIndex].Replace("\"","")
+
+ //if i%50=0 then printfn "%04i/%i: %s" i rowCount identifier
+ //printfn "%04i/%i: %s" i rowCount identifier
+ let annotations =
+ if identifier = "" then
+ Array.init annotations.Length (fun _ -> "")
+ else
+ Mapping.getAnnotationRow annotationFrame annotations multipleIdentifierSeparator multipleAnnotationSeparator truncateID identifier
+ Seq.append x annotations
+ |> String.concat columnSeparator
+ )
+ |> Array.append [|header|]
+
+ let annotateAndWriteData annotationFrame inputPath (columnSeparator:string) columnHeader truncateID (annotations:string[]) (multipleIdentifierSeparator:string) (multipleAnnotationSeparator:string) outputPath =
+ /// warns user if separators are identical
+ let checkSeparators =
+ if columnSeparator = multipleIdentifierSeparator then failwith "WARNING: Column separator is equal to identifier separator"
+ if columnSeparator = multipleAnnotationSeparator then failwith "WARNING: Column separator is equal to annotation separator"
+ let annotatedRows =
+ getAnnotatedLines annotationFrame inputPath columnSeparator columnHeader annotations multipleIdentifierSeparator truncateID multipleAnnotationSeparator
+ System.IO.File.WriteAllLines(outputPath,annotatedRows)
+
+module SamPlotting =
+ open Deedle
+ open FSharp.Stats
+ open Plotly.NET
+ open FSharp.Stats.Testing
+ open SAM
+ open Plotly.NET.StyleParam
+
+ let createSAMChart res =
+
+ let observed = [| res.NegSigBioitem; res.NonSigBioitem; res.PosSigBioitem|] |> Array.concat
+ let obs = observed |> Array.map (fun x -> x.Statistics)
+ let expected = res.AveragePermutations |> Array.map (fun x -> x.Statistics)
+ let minDi = Seq.min obs
+ let maxDi = Seq.max obs
+
+
+ // positive significant changes
+ let posExpected = expected.[res.NegSigBioitem.Length + res.NonSigBioitem.Length .. res.NegSigBioitem.Length + res.NonSigBioitem.Length + res.PosSigBioitem.Length-1]
+ let posChart =
+ Chart.Point(posExpected,res.PosSigBioitem |> Array.map (fun x -> x.Statistics))
+ |> Chart.withLineStyle(Color=Color.fromKeyword Green)
+ |> Chart.withTraceInfo("positive change",Visible = Visible.True )
+
+
+ // no significant changes
+ let nonex = expected.[res.NegSigBioitem.Length .. res.NegSigBioitem.Length + res.NonSigBioitem.Length-1]
+ let nonchart =
+ Chart.Point(nonex,res.NonSigBioitem |> Array.map (fun x -> x.Statistics))
+ |> Chart.withLineStyle(Color=Color.fromKeyword Gray)
+ |> Chart.withTraceInfo("no change",Visible = Visible.True)
+
+ // negative significant changes
+ let negex = expected.[0 .. res.NegSigBioitem.Length-1]
+ let negchart =
+ Chart.Point(negex,res.NegSigBioitem |> Array.map (fun x -> x.Statistics))
+ |> Chart.withLineStyle(Color=Color.fromKeyword Red)
+ |> Chart.withTraceInfo("negative change",Visible = Visible.True)
+
+ let samValues =
+ [
+ negchart
+ nonchart
+ posChart
+ ]
+ |> Chart.combine
+
+ let chartConfig =
+ let svdConfig =
+ ConfigObjects.ToImageButtonOptions.init(
+ Format = StyleParam.ImageFormat.SVG)
+ Config.init (
+ ToImageButtonOptions = svdConfig,
+ ModeBarButtonsToAdd=[ModeBarButton.HoverCompareCartesian]
+
+ )
+
+ let cutLineUp = [(minDi + res.Delta) ; (maxDi + res.Delta)]
+ let cutsUp =
+ Chart.Line(cutLineUp,[minDi;maxDi])
+ |> Chart.withLineStyle(Color=Color.fromKeyword Purple,Dash = StyleParam.DrawingStyle.Dash, Width = 0.5)
+ |> Chart.withTraceInfo("delta",Visible = Visible.True)
+ let cutLineLow = [(minDi - res.Delta) ; (maxDi - res.Delta)]
+ let cutsLow =
+ Chart.Line(cutLineLow,[minDi;maxDi])
+ |> Chart.withLineStyle(Color=Color.fromKeyword Purple,Dash = StyleParam.DrawingStyle.Dash, Width = 0.5)
+ |> Chart.withTraceInfo("delta",Visible = Visible.True)
+ let linechart =
+ Chart.Line([minDi;maxDi], [minDi;maxDi])
+ |> Chart.withTraceInfo("bisecting angle",Visible = Visible.True)
+ |> Chart.withLineStyle(Color=Color.fromKeyword Black, Width = 1)
+
+ let uppercut =
+ let xAnchorUppercut = [minDi .. 5. .. maxDi]
+ Chart.Line (xAnchorUppercut, List.init xAnchorUppercut.Length (fun x -> res.UpperCut))
+
+ |> Chart.withTraceInfo("upper cut",Visible = Visible.True)
+ |> Chart.withLineStyle(Color=Color.fromKeyword Black,Dash = StyleParam.DrawingStyle.Dash, Width = 0.3)
+
+ let lowercut =
+ Chart.Line([minDi;maxDi],[res.LowerCut;res.LowerCut])
+ |> Chart.withTraceInfo("lower cut",Visible = Visible.True)
+ |> Chart.withLineStyle(Color=Color.fromKeyword Black,Dash = StyleParam.DrawingStyle.Dash,Width = 0.3)
+ //|> Chart.withXAxisStyle(MinMax = (-15,20))
+
+ |> Chart.withTraceInfo("lower cut",Visible = Visible.True)
+ let plot =
+ [linechart;
+ samValues;
+ cutsUp;
+ cutsLow;
+ uppercut;
+ lowercut]
+ |> Chart.combine
+ |> Chart.withTitle(title = "SAM results")
+ |> Chart.withXAxisStyle("expected Score")
+ |> Chart.withYAxisStyle ("observed Score")
+ |> Chart.withConfig(chartConfig)
+ |> Chart.withTemplate(ChartTemplates.lightMirrored)
+ plot
+
+module Plotting =
+
+ open FSharp.Stats
+ open Plotly.NET
+ open Deedle
+ open BioFSharp.IO
+ open DeedleExtensions
+
+ let plotMA highlightList (f:Frame<string,string>) =
+ let dataAll =
+ f
+ |> Frame.mapRows (fun k s ->
+ let a = s.GetAs<float>("MeanAbundance",nan)
+ let r = s.GetAs<float>("MeanR",nan)
+ let stDev = s.GetAs<float>("StDevR",nan)
+ let l =
+ let x = s.GetAs<string>("ProtName","")
+ if x = "" then
+ ""
+ else
+ x
+ let c = s.GetAs<bool>("IsCandidate")
+ {|
+ ProtName = l
+ MeanAbundance = a
+ MeanR = r
+ StDevR = stDev
+ IsCandidate = c
+ |}
+ )
+ |> Series.values
+ |> Array.ofSeq
+
+ let all =
+ let x = dataAll |> Seq.map (fun x -> x.MeanR)
+ let y = dataAll |> Seq.map (fun x -> x.MeanAbundance)
+ let label = dataAll |> Seq.map (fun x -> x.ProtName)
+ Chart.Point(y,x, MultiText = label)
+
+ let candidates =
+ let dataFiltered = dataAll |> Array.filter (fun x -> x.IsCandidate)
+ let x = dataFiltered |> Seq.map (fun x -> x.MeanR)
+ let y = dataFiltered |> Seq.map (fun x -> x.MeanAbundance)
+ let labels = dataFiltered |> Seq.map (fun x -> x.ProtName)
+ Chart.Point(
+ y,x,
+ Name="points",
+ MultiText=labels,
+ TextPosition=StyleParam.TextPosition.TopRight
+ )
+ |> Chart.withTraceInfo "candidates"
+ let candidatesNoStabw =
+ let dataFiltered = dataAll |> Array.filter (fun x -> x.IsCandidate)
+ let x = dataFiltered |> Seq.map (fun x -> x.MeanR)
+ let y = dataFiltered |> Seq.map (fun x -> x.MeanAbundance)
+ let labels = dataFiltered |> Seq.map (fun x -> x.ProtName)
+
+ Chart.Point(
+ y,x,
+ Name="points", MultiText = labels
+ )
+ |> Chart.withTraceInfo "candidatesNoStabw"
+ let candidates' =
+ let dataFiltered = dataAll |> Array.filter (fun x -> highlightList |> List.contains x.ProtName )
+ let x = dataFiltered |> Seq.map (fun x -> x.MeanR)
+ let y = dataFiltered |> Seq.map (fun x -> x.MeanAbundance)
+ let labels = dataFiltered |> Seq.map (fun x -> x.ProtName)
+ Chart.Point(
+ y,x,
+ Name="points",
+ MultiText=labels,
+ TextPosition=StyleParam.TextPosition.TopRight
+ )
+ |> Chart.withTraceInfo "candidates Reci"
+ [
+ all
+ candidatesNoStabw
+ candidates'
+
+ ]
+ |> Chart.combine
+ |> Chart.withYAxisStyle "log2(bait)-log2(control)"
+ |> Chart.withXAxisStyle "log2(Mean intensity)"
+
+ let plotVulcano highlightList f =
+ let dataAll =
+ f
+ |> Frame.mapRows (fun k s ->
+ let a = s.GetAs<float>("MeanAbundance",nan)
+ let r = s.GetAs<float>("MeanR",nan)
+ let stDev = s.GetAs<float>("StDevR",nan)
+ let qValue = s.GetAs<float>("qValue",nan)
+ let l =
+ let x = s.GetAs<string>("ProtName","")
+ if x = "" then
+ ""
+ else
+ x
+ let c = s.GetAs<bool>("IsCandidate")
+ {|
+ ProtName = l
+ MeanAbundance = a
+ MeanR = r
+ StDevR = stDev
+ IsCandidate = c
+ QVal = qValue
+ |}
+ )
+ |> Series.values
+ |> Array.ofSeq
+
+ let all =
+ let y = dataAll |> Seq.map (fun x -> x.MeanR)
+ let x =
+ dataAll
+ |> Seq.map (fun x -> x.QVal)
+ |> Seq.map (fun x ->
+ let res = -log10 x
+ if infinity.Equals res then 4. else res
+ )
+ let label = dataAll |> Seq.map (fun x -> x.ProtName)
+ let xError =
+ dataAll |> Seq.map (fun x -> x.StDevR)
+ Chart.Point(y,x, MultiText = label)
+ let candidatesNoStabw =
+ let dataFiltered = dataAll |> Array.filter (fun x -> x.IsCandidate)
+ let y = dataFiltered |> Seq.map (fun x -> x.MeanR)
+ let x =
+ dataFiltered
+ |> Seq.map (fun x -> x.QVal)
+ |> Seq.map (fun x ->
+ let res = -log10 x
+ if infinity.Equals res then 4. else res
+ )
+ let labels = dataFiltered |> Seq.map (fun x -> x.ProtName)
+ let xError = dataFiltered |> Seq.map (fun x -> x.StDevR)
+ Chart.Point(
+ y,x,
+ Name="points",
+ MultiText=labels
+ )
+ |> Chart.withTraceInfo "candidatesNoStabw"
+ let candidates' =
+ let dataFiltered = dataAll |> Array.filter (fun x -> highlightList |> List.contains x.ProtName )
+ let y = dataFiltered |> Seq.map (fun x -> x.MeanR)
+ let x =
+ dataFiltered
+ |> Seq.map (fun x -> x.QVal)
+ |> Seq.map (fun x ->
+ let res = -log10 x
+ if infinity.Equals res then 4. else res
+ )
+ let labels = dataFiltered |> Seq.map (fun x -> x.ProtName)
+ Chart.Point(
+ y,x,
+ Name="points",
+ MultiText=labels,
+ TextPosition=StyleParam.TextPosition.TopRight
+ )
+ |> Chart.withTraceInfo "candidates Reci"
+ [
+ all
+ candidatesNoStabw
+ candidates'
+
+ ]
+ |> Chart.combine
+ |> Chart.withYAxisStyle "log2(bait)-log2(control)"
+ |> Chart.withXAxisStyle "log2(Mean intensity)"
+
+ let plotHisto (f:Frame<string,string>) =
+ let dataAll =
+ f
+ |> Frame.mapRows (fun k s ->
+ let a = s.GetAs<float>("MeanAbundance",nan)
+ let r = s.GetAs<float>("MeanR",nan)
+ let stDev = s.GetAs<float>("StDevR",nan)
+ let l = ""//s.GetAs<string>("ProtName","")
+ let c = s.GetAs<bool>("IsCandidate")
+ {|
+ ProtName = l
+ MeanAbundance = a
+ MeanR = r
+ StDevR = stDev
+ IsCandidate = c
+ |}
+ )
+ |> Series.values
+ |> Array.ofSeq
+ let histo =
+ let data = dataAll |> Seq.map (fun x -> x.MeanR) |> Seq.filter (fun x -> nan.Equals x |> not)|> Array.ofSeq
+ let bw = FSharp.Stats.Distributions.Bandwidth.nrd0 data
+ let binned = FSharp.Stats.Distributions.Frequency.create (bw) data |> Map.toSeq
+ Chart.Column binned
+ |> Chart.withXAxisStyle "log2(bait)-log2(control)"
+ histo
+
+module MetaxaTemp =
+ open MetaXa
+ open MetaFrame
+ open Stats
+ open FSharp.Stats
+ open Plotly.NET
+ open Deedle
+ open BioFSharp.IO
+ open DeedleExtensions
+ open ISADotNet
+
+
+ module Global =
+
+ /// As used by Deseq2, see: https://github.com/hbctraining/DGE_workshop/blob/master/lessons/02_DGE_count_normalization.md
+ module MedianOfRatios =
+
+ /// expects data frame with shape frame<sampleIds,GeneIds>
+ let medianOfRatiosWide (data:Frame<string,string>) =
+ //data frame with shape frame<sampleIds,GeneIds>
+ let ratios =
+ data
+ |> Frame.dropSparseCols
+ |> Frame.getNumericCols
+ |> Series.map (fun geneID s ->
+ let geometricMeanAcrossAllSamples =
+ s.Values
+ |> FSharp.Stats.Seq.median
+ let ratioPerSample =
+ s
+ |> Series.map (fun x sample -> sample / geometricMeanAcrossAllSamples)
+ ratioPerSample
+ )
+ |> Frame.ofColumns
+ // Series of shape: Series<sampleId,correctionFactor:float>
+ let correctionFactorPerSample =
+ ratios
+ |> Frame.transpose
+ |> Stats.median
+ // data frame with frame<sampleIds,GeneIds>
+ // contains gene data divided by the sample wise estimated correction factor
+ let correctedData =
+ data
+ |> Frame.getNumericCols
+ |> Series.map (fun geneID s ->
+ s / correctionFactorPerSample
+ )
+ |> Frame.ofColumns
+ let abundanceInit =
+ Chart.Column(data |> Frame.transpose |> Frame.dropSparseRows |> Stats.median |> Series.observations)
+ |> Chart.withXAxisStyle "SampleId"
+ |> Chart.withXAxisStyle "Median of Initial Abundances"
+ let c =
+ Chart.Column(correctionFactorPerSample |> Series.observations)
+ |> Chart.withXAxisStyle "SampleId"
+ |> Chart.withXAxisStyle "Correction Factor."
+ let abundanceCorrected =
+ Chart.Column(correctedData |> Frame.transpose |> Frame.dropSparseRows |> Stats.median |> Series.observations)
+ |> Chart.withXAxisStyle "SampleId"
+ |> Chart.withXAxisStyle "Median of corrected abundances"
+ correctedData,abundanceInit,c,abundanceCorrected
+
+
+ module Imputation =
+
+
+ module SimpleSampling =
+ /// Expects normalized data.
+ /// (learnVarianceMeanDependence: seq<seq<float>> -> (float -> float)) function, which learns how to map from a mean expression to an estimator for variance.
+ let computeColumnWiseUsing (sampleFunction: float -> float -> float) minCountsForVarEst (quantileForLowMean:float option) (learnVarianceMeanDependence: seq<float*float> -> (float -> float)) (grouped:Frame<string*string,string>) =
+ ///
+ let meanWithAtLeast k (x:Series<_,float>) =
+ let tmp = Series.values x
+ if tmp |> Seq.length < k then
+ nan
+ else
+ x |> Series.values |> Seq.mean
+ ///
+ let cvWithAtLeast k (x:Series<_,float>) =
+ let tmp = Series.values x
+ if tmp |> Seq.length < k then
+ nan
+ else
+ x |> Series.values |> cv //Seq.disp
+
+ /// Computes means and variances of complete replicates.
+ let log2MeanToCVEmpirical =
+ grouped
+ |> Frame.getNumericCols
+ |> Series.map (fun geneID s ->
+ let means = s |> (meanWithAtLeast minCountsForVarEst)
+ let cols = s |> (cvWithAtLeast minCountsForVarEst)
+ means, cols
+ )
+ |> Series.values
+ |> Array.ofSeq
+ |> Array.filter (fun x -> nan.Equals(fst x) |> not && nan.Equals(snd x) |> not)
+ |> Array.sortBy fst
+ |> Array.map (fun (mean,dis) -> log2 mean, dis)
+ |> fun data ->
+ let borders = FSharp.Stats.Testing.Outliers.tukey 1.5 (data |> Array.map snd)
+ data
+ |> Array.filter (fun x -> snd x > borders.Lower && snd x < borders.Upper )
+ /// function mapping from a mean expression to an estimator for dispersion (cv).
+ let log2MeanToCv =
+ learnVarianceMeanDependence log2MeanToCVEmpirical
+
+ let log2MeanToCVByFit =
+ log2MeanToCVEmpirical
+ |> Array.map (fun x -> fst x, fst x |> log2MeanToCv)
+ let meanToCvEstChart =
+ [
+ [
+
+ Chart.PointDensity(log2MeanToCVEmpirical |> Array.map fst,log2MeanToCVEmpirical |> Array.map snd)
+ // |> Chart.withMarkerStyle (3)
+ ]
+ |> Chart.combine
+ |> Chart.withTraceInfo "mean to CV empirical"
+
+ Chart.Spline(log2MeanToCVByFit)
+ |> Chart.withTraceInfo "mean Vs. CV fitted"
+ ]
+ |> Chart.combine
+ |> Chart.withXAxisStyle "log2(mean Abundance)"
+ |> Chart.withYAxisStyle "CV"
+
+ let quantileForLowMean, quantilesToMeansChart =
+ let data =
+ Array.map fst log2MeanToCVEmpirical
+ |> Array.map (fun x -> 2.**x)
+ match quantileForLowMean with
+ | None ->
+ None,
+ [
+ [|0.05 .. 0.05 .. 0.95|]
+ |> Array.map (fun q -> q,FSharp.Stats.Quantile.compute q data)
+ |> Chart.Point
+ |> Chart.withTraceInfo "quantiles VS Abundance"
+ ]
+ |> Chart.combine
+ |> Chart.withXAxisStyle "Quantiles"
+ |> Chart.withYAxisStyle "Abundance"
+ | Some quantileForLowMean ->
+ let q = data |> FSharp.Stats.Quantile.compute quantileForLowMean
+ Some q,
+ [
+ [|0.00 .. 0.05 .. 0.95|]
+ |> Array.mapi (fun i q -> if i = 0 then 0.01 else q)
+ |> Array.map (fun q -> q,FSharp.Stats.Quantile.compute q data)
+ |> Chart.Point
+ |> Chart.withTraceInfo "quantiles VS Abundance"
+ Chart.Point([quantileForLowMean],[q])
+ |> Chart.withTraceInfo "selected quantile"
+ ]
+ |> Chart.combine
+ |> Chart.withXAxisStyle "Quantiles"
+ |> Chart.withYAxisStyle "Abundance"
+
+ /// Imputation by sampling from a gausian normal distribution based on the input vector
+ /// function is truncated at zero. Can not return negative values.
+ let initImputeUsing sampleF (estimateCv:float -> float) (lowMean:float option) (d: seq<float>) =
+ let clean = d |> Seq.filter (fun x -> nan.Equals x |> not)
+ let sampleSafe mean standardDev =
+ let rec sample () =
+ let tmp = sampleF mean standardDev
+ if tmp < 0. then sample () else tmp
+ sample()
+ if clean |> Seq.isEmpty |> not then
+ let mean = Seq.mean clean
+ let cvEst = estimateCv mean
+ let standardDev = cvEst * mean
+ //printfn "mean %f" mean
+ //printfn "stndev %f" standardDev
+ //printfn "before %A" d
+ let res =
+ d
+ |> Seq.map (fun x -> if nan.Equals x then sampleSafe mean standardDev else x)
+ //printfn "%A" res
+ res
+ else
+ match lowMean with
+ | Some lowMean ->
+ let mean = lowMean
+ let cvEst = estimateCv mean
+ let standardDev = cvEst * mean
+ d
+ |> Seq.map (fun x -> if nan.Equals x then sampleSafe mean standardDev else x)
+ | None -> d
+
+ let imputeUsing (x:seq<float>) =
+ initImputeUsing sampleFunction (log2 >> log2MeanToCv) (quantileForLowMean) x
+
+ /// Computes means and variances of complete replicates.
+ let imputed =
+ // Step 1: Set genes missing when they are not comparable
+ // because all sample groups have 1 or zero observations.
+ grouped
+ |> Frame.getNumericCols
+ |> Series.map (fun geneID s ->
+ let geneDataByFactorLevels =
+ s
+ |> Series.applyLevel fst (Series.values >> Array.ofSeq)
+ let geneData =
+ geneDataByFactorLevels
+ |> Series.observations
+ let present =
+ geneData
+ |> Seq.map (fun (sampleID,v) ->
+ v.Length > 1
+ )
+ |> Seq.contains true
+ if present then s else s |> Series.mapValues (fun s -> nan)
+ )
+ |> Frame.ofColumns
+ // Step 2:
+ // x present 0 not present *nan*
+ // [x; x; x] -> [x; x; x]
+ // [x; x; 0] -> [x; x; x] by sampling, mean from sample, var by est
+ // Will be biased towards the remaining point.
+ // [x; 0; 0] -> [x; x; x] by sampling, mean from sample, var by est
+ //
+ // [0; 0; 0] ->
+ //depending on the value of quantileForLowMean :
+ //Some x -> [x; x; x]
+ //None -> [0; 0; 0]
+ |> Map.mapColWiseGroupedByFactorLevels imputeUsing
+
+ // meanToCvEstChart |> Chart.show
+ // quantilesToMeansChart |> Chart.show
+ imputed
+
open Annotate
open Argu
-#load "scripts/SamPlotting.fsx"
open SamPlotting
-#load "scripts/Plotting.fsx"
open Plotting
-#load "scripts/MetaxaTemp.fsx"
open MetaxaTemp
module Arguments =
type WorkflowArgs =
| [<Unique>] [<Mandatory>] [<AltCommandLine("-i")>] InputPath of path:string
+ | [<Unique>] [<Mandatory>] [<AltCommandLine("-a")>] ArcPath of path:string
+ | [<Unique>] [<Mandatory>] [<AltCommandLine("-m")>] MappingFilePath of path:string
+ | [<Unique>] [<Mandatory>] [<AltCommandLine("-id")>] RunID of path:string
// | [<Unique>] [<Mandatory>] [<AltCommandLine("-o")>] OutputPath of path:string
| [<Unique>] [<AltCommandLine("-c")>] IdentifierColumnHeader of string
| [<Unique>] [<AltCommandLine("-fdr")>] FDR of string
@@ -43,6 +684,9 @@ module Arguments =
member this.Usage =
match this with
| InputPath x -> "relative input file path in /arc/"
+ | ArcPath x -> "relative path to the arc"
+ | RunID x -> "ID for the run"
+ | MappingFilePath x -> "relative path to the mapping file path in /arc/"
// | OutputPath x -> "relative output file path in /arc/runs/"
| IdentifierColumnHeader x -> "column header of identifier column file to annotate"
| FDR x -> "FDR"
@@ -61,6 +705,9 @@ module Arguments =
// arguments are converted
let inputPath = annotationR.GetResult(InputPath)//@"/arc/assays/" + annotationR.GetResult(InputPath)
+ let arcPath = annotationR.GetResult(ArcPath)
+ let runID = annotationR.GetResult(RunID)
+ let mappingFilePath = annotationR.GetResult(MappingFilePath)
// let outputFilePath = annotationR.GetResult(OutputPath)
let columnHeader =
match annotationR.TryGetResult(IdentifierColumnHeader) with
@@ -90,10 +737,8 @@ module Arguments =
////////////////////////////////////////
////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////
-let runID = System.Guid.NewGuid().ToString()
-let source = __SOURCE_DIRECTORY__
-let runP = System.IO.Path.Combine([|source + @"\..\..\runs";runID|])
-System.IO.Directory.CreateDirectory runP
+let runID = Arguments.runID
+System.IO.Directory.CreateDirectory runID
|> ignore
////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////
@@ -102,10 +747,9 @@ System.IO.Directory.CreateDirectory runP
////////////////////////////////////////
////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////
-let arcPath = __SOURCE_DIRECTORY__ + @"\..\..\"
let qi =
- arcIO.NET.Investigation.fromArcFolder arcPath
+ arcIO.NET.Investigation.fromArcFolder Arguments.arcPath
|> QueryModel.QInvestigation.fromInvestigation
// qi.ProtocolNames
@@ -151,10 +795,10 @@ let qSampleNameToIntensityColumn fN =
////////////////////////////////////////
////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////
-let mappingFrame = Mapping.readMapping (__SOURCE_DIRECTORY__ + @"\scripts\chlamy_jgi55.txt") "\t" "Identifier"
+let mappingFrame = Mapping.readMapping Arguments.mappingFilePath "\t" "Identifier"
mappingFrame.Print()
let truncID (id : string) = (id.Split([|".p";".t"|],System.StringSplitOptions.None)).[0]
-let outPathA = runP + "/AnnotatedResult.tsv"
+let outPathA = runID + "/AnnotatedResult.tsv"
Data.annotateAndWriteData
mappingFrame
@@ -525,45 +1169,45 @@ let doAnalysis (allDataForAnalysis:Frame<string,string>) dataForAnalysis (allCon
/////////////////////////////////////////////////////
// Save Results
/////////////////////////////////////////////////////
- let outP = System.IO.Path.Combine([|runP;baitGroup|])
+ let outP = System.IO.Path.Combine([|runID;baitGroup|])
System.IO.Directory.CreateDirectory outP
|> ignore
abundanceChart
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\1_Norm_PreNormGlobalAbundanceChart.html")
+ |> Chart.saveHtml (outP + @"/1_Norm_PreNormGlobalAbundanceChart.html")
corrFChart
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\2_Norm_CorrectionfactorChart.html")
+ |> Chart.saveHtml (outP + @"/2_Norm_CorrectionfactorChart.html")
corrAbundanceChart
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\3_Norm_AfterNormGlobalAbundanceChart.html")
+ |> Chart.saveHtml (outP + @"/3_Norm_AfterNormGlobalAbundanceChart.html")
preImpChart
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\4_Imputation_preImpChart.html")
+ |> Chart.saveHtml (outP + @"/4_Imputation_preImpChart.html")
afterImpChart
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\5_Imputation_afterImpChart.html")
+ |> Chart.saveHtml (outP + @"/5_Imputation_afterImpChart.html")
preNormChart
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\6_QuantilNorm_PreNormChart.html")
+ |> Chart.saveHtml (outP + @"/6_QuantilNorm_PreNormChart.html")
afterNormChart
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\7_QuantilNorm_AfterNormChart.html")
+ |> Chart.saveHtml (outP + @"/7_QuantilNorm_AfterNormChart.html")
createSAMChart res
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\8_Testing_SAM.html")
+ |> Chart.saveHtml (outP + @"/8_Testing_SAM.html")
plotHisto finPlot
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\9_Testing_DifferencesHistogram.html")
+ |> Chart.saveHtml (outP + @"/9_Testing_DifferencesHistogram.html")
plotMA [] finPlot // ["CDJ5";"LON";"PSBQLD";"PGRL1"]
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\10_Testing_MAPlot.html")
+ |> Chart.saveHtml (outP + @"/10_Testing_MAPlot.html")
plotVulcano [] finPlot // ["CDJ5";"LON";"PSBQLD";"PGRL1"]
|> Chart.withTemplate ChartTemplates.lightMirrored
- |> Chart.saveHtml (outP + @"\11_Testing_VulcanoPlot.html")
- toSave.SaveCsv(outP + @"\Analysis_complete.tsv",includeRowKeys=true,keyNames=["Protein"],separator='\t')
- System.IO.File.WriteAllLines(outP + @"\params.txt",fsi.CommandLineArgs)
+ |> Chart.saveHtml (outP + @"/11_Testing_VulcanoPlot.html")
+ toSave.SaveCsv(outP + @"/Analysis_complete.tsv",includeRowKeys=true,keyNames=["Protein"],separator='\t')
+ System.IO.File.WriteAllLines(outP + @"/params.txt",fsi.CommandLineArgs)
printfn "Finished Analysis! results can be found at:%s" outP
////////////////////////////////////////////////////////////////////////////////
diff --git a/workflows/EvalTurboID/TurboIDSampleArc.cwl b/workflows/EvalTurboID/TurboIDSampleArc.cwl
new file mode 100644
index 0000000..e3fbda7
--- /dev/null
+++ b/workflows/EvalTurboID/TurboIDSampleArc.cwl
@@ -0,0 +1,48 @@
+cwlVersion: v1.2
+class: CommandLineTool
+hints:
+ DockerRequirement:
+ dockerPull: mcr.microsoft.com/dotnet/sdk:7.0
+requirements:
+ # - class: InlineJavascriptRequirement
+ # - class: InitialWorkDirRequirement
+ # listing:
+ # - entry: $(inputs.scriptDirectory)
+ # writable: true
+ - class: EnvVarRequirement
+ envDef:
+ - envName: DOTNET_NOLOGO
+ envValue: "true"
+ - class: NetworkAccess
+ networkAccess: true
+baseCommand: [dotnet, fsi]
+inputs:
+ scriptFile:
+ type: File
+ inputBinding:
+ position: 1
+ input:
+ type: File
+ inputBinding:
+ position: 2
+ prefix: -i
+ mappingFile:
+ type: File
+ inputBinding:
+ position: 3
+ prefix: -m
+ arcDirectory:
+ type: Directory
+ inputBinding:
+ position: 4
+ prefix: -a
+ runID:
+ type: string
+ inputBinding:
+ position: 5
+ prefix: -id
+outputs:
+ output:
+ type: Directory
+ outputBinding:
+ glob: $(runtime.outdir)/$(inputs.runID)
\ No newline at end of file
--
GitLab