Merge remote-tracking branch 'upstream/master'

trishorts · Nov 5, 2024 · e9d14c8 · e9d14c8
2 parents c9551e3 + e5cf73e
commit e9d14c8
Show file tree

Hide file tree

Showing 165 changed files with 10,132 additions and 930 deletions.
diff --git a/.github/workflows/dotnet.yml b/.github/workflows/dotnet.yml
@@ -15,7 +15,7 @@ jobs:
     - name: Set up .NET
       uses: actions/setup-dotnet@v1
       with:
-        dotnet-version: 6.0.x
+        dotnet-version: 8.0.x
     - name: Restore dependencies
       run: cd mzLib && dotnet restore
     - name: Build
@@ -25,9 +25,9 @@ jobs:
     - name: Build (TestFlashLFQ)
       run: cd mzLib && dotnet build --no-restore ./TestFlashLFQ/TestFlashLFQ.csproj
     - name: Add coverlet collector (Test)
-      run: cd mzLib && dotnet add Test/Test.csproj package coverlet.collector -v 6.0.0
+      run: cd mzLib && dotnet add Test/Test.csproj package coverlet.collector -v 6.0.2
     - name: Add coverlet collector (TestFlashLFQ)
-      run: cd mzLib && dotnet add TestFlashLFQ/TestFlashLFQ.csproj package coverlet.collector -v 6.0.0
+      run: cd mzLib && dotnet add TestFlashLFQ/TestFlashLFQ.csproj package coverlet.collector -v 6.0.2
     - name: Test
       run: cd mzLib && dotnet test --no-build --verbosity normal --collect:"XPlat Code Coverage" /p:CoverletOutputFormat=cobertura ./Test/Test.csproj
     - name: TestFlashLFQ

diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml
@@ -14,7 +14,7 @@ jobs:
       - name: Setup .NET
         uses: actions/setup-dotnet@v1
         with:
-          dotnet-version: 6.0.x
+          dotnet-version: 8.0.x
       - name: Set up NuGet
         uses: nuget/setup-nuget@v1
         with:

diff --git a/mzLib/BayesianEstimation/BayesianEstimation.csproj b/mzLib/BayesianEstimation/BayesianEstimation.csproj
@@ -1,7 +1,7 @@
 <Project Sdk="Microsoft.NET.Sdk">
 
   <PropertyGroup>
-    <TargetFramework>net6.0</TargetFramework>
+    <TargetFramework>net8.0</TargetFramework>
     <Platforms>x64</Platforms>
   </PropertyGroup>
 
@@ -11,6 +11,7 @@
   </PropertyGroup>
 
   <ItemGroup>
+    <PackageReference Include="CsvHelper" Version="32.0.3" />
     <PackageReference Include="MathNet.Numerics" Version="5.0.0" />
   </ItemGroup>
 

diff --git a/mzLib/Chemistry/ChemicalFormula.cs b/mzLib/Chemistry/ChemicalFormula.cs
@@ -19,9 +19,11 @@
 using MzLibUtil;
 using System;
 using System.Collections.Generic;
+using System.ComponentModel.DataAnnotations.Schema;
 using System.Globalization;
 using System.Linq;
 using System.Text;
+using System.Text.Json.Serialization;
 using System.Text.RegularExpressions;
 
 namespace Chemistry
@@ -65,7 +67,7 @@ public ChemicalFormula(IHasChemicalFormula capFormula)
             Elements = new Dictionary<Element, int>(capFormula.ThisChemicalFormula.Elements);
         }
 
-        public ChemicalFormula ThisChemicalFormula => this;
+        [JsonIgnore] public ChemicalFormula ThisChemicalFormula => this;
 
         /// <summary>
         /// Gets the average mass of this chemical formula

diff --git a/mzLib/Chemistry/Chemistry.csproj b/mzLib/Chemistry/Chemistry.csproj
@@ -1,14 +1,18 @@
 <Project Sdk="Microsoft.NET.Sdk">
 
   <PropertyGroup>
-    <TargetFramework>net6.0</TargetFramework>
+    <TargetFramework>net8.0</TargetFramework>
     <Platforms>x64</Platforms>
   </PropertyGroup>
 
   <PropertyGroup Condition="'$(Configuration)'=='Release'">
     <DebugType>full</DebugType>
     <DebugSymbols>true</DebugSymbols>
   </PropertyGroup>
+
+  <ItemGroup>
+    <PackageReference Include="CsvHelper" Version="32.0.3" />
+  </ItemGroup>
 
   <ItemGroup>
     <ProjectReference Include="..\MzLibUtil\MzLibUtil.csproj" />

diff --git a/mzLib/Chemistry/ClassExtensions.cs b/mzLib/Chemistry/ClassExtensions.cs
@@ -48,6 +48,7 @@ public static double ToMass(this double massToChargeRatio, int charge)
             return Math.Abs(charge) * massToChargeRatio - charge * Constants.ProtonMass;
         }
 
+        public static double? RoundedDouble(this double myNumber, int places = 9) => RoundedDouble(myNumber as double?, places);
         public static double? RoundedDouble(this double? myNumber, int places = 9)
         {
             if (myNumber != null)

diff --git a/mzLib/Development/Deconvolution/TestCases/TestDevelopmentTestCases.cs b/mzLib/Development/Deconvolution/TestCases/TestDevelopmentTestCases.cs
@@ -1,5 +1,6 @@
 using MzLibUtil;
 using NUnit.Framework;
+using Assert = NUnit.Framework.Legacy.ClassicAssert;
 using MassSpectrometry;
 using System.Diagnostics.CodeAnalysis;
 using Readers;

diff --git a/mzLib/Development/Development.csproj b/mzLib/Development/Development.csproj
@@ -1,16 +1,17 @@
 <Project Sdk="Microsoft.NET.Sdk">
 
   <PropertyGroup>
-    <TargetFramework>net6.0</TargetFramework>
+    <TargetFramework>net8.0</TargetFramework>
 	  <Platforms>x64</Platforms>
     <ImplicitUsings>enable</ImplicitUsings>
     <Nullable>enable</Nullable>
   </PropertyGroup>
 
   <ItemGroup>
-    <PackageReference Include="Microsoft.NET.Test.Sdk" Version="17.1.0" />
-    <PackageReference Include="NUnit" Version="3.13.3" />
-    <PackageReference Include="NUnit3TestAdapter" Version="4.2.1" />
+    <PackageReference Include="CsvHelper" Version="32.0.3" />
+    <PackageReference Include="Microsoft.NET.Test.Sdk" Version="17.10.0" />
+    <PackageReference Include="NUnit" Version="4.1.0" />
+    <PackageReference Include="NUnit3TestAdapter" Version="4.5.0" />
   </ItemGroup>
 
   <ItemGroup>

diff --git a/mzLib/FlashLFQ/ChromatographicPeak.cs b/mzLib/FlashLFQ/ChromatographicPeak.cs
@@ -1,22 +1,38 @@
-using Chemistry;
-using MathNet.Numerics.Statistics;
-using System;
+using System;
 using System.Collections.Generic;
 using System.Linq;
 using System.Text;
+using ClassExtensions = Chemistry.ClassExtensions;
+using FlashLFQ.PEP;
 
 namespace FlashLFQ
 {
     public class ChromatographicPeak
     {
         public double Intensity;
+        public double ApexRetentionTime => Apex?.IndexedPeak.RetentionTime ?? -1;
         public readonly SpectraFileInfo SpectraFileInfo;
         public List<IsotopicEnvelope> IsotopicEnvelopes;
+        public int ScanCount => IsotopicEnvelopes.Count;
         public double SplitRT;
         public readonly bool IsMbrPeak;
         public double MbrScore;
+        public double PpmScore { get; set; }
+        public double IntensityScore { get; set; }
+        public double RtScore { get; set; }
+        public double ScanCountScore { get; set; }
+        public double IsotopicDistributionScore { get; set; }
+        /// <summary>
+        /// Stores the pearson correlation between the apex isotopic envelope and the theoretical isotopic distribution
+        /// </summary>
+        public double IsotopicPearsonCorrelation => Apex?.PearsonCorrelation ?? -1;
+        public double RtPredictionError { get; set; }
+        public List<int> ChargeList { get; set; }
+        internal double MbrQValue { get; set; }
+        public ChromatographicPeakData PepPeakData { get; set; }
+        public double? MbrPep { get; set; }
 
-        public ChromatographicPeak(Identification id, bool isMbrPeak, SpectraFileInfo fileInfo)
+        public ChromatographicPeak(Identification id, bool isMbrPeak, SpectraFileInfo fileInfo, bool randomRt = false)
         {
             SplitRT = 0;
             NumChargeStatesObserved = 0;
@@ -27,6 +43,14 @@ public ChromatographicPeak(Identification id, bool isMbrPeak, SpectraFileInfo fi
             IsotopicEnvelopes = new List<IsotopicEnvelope>();
             IsMbrPeak = isMbrPeak;
             SpectraFileInfo = fileInfo;
+            RandomRt = randomRt;
+        }
+
+        public bool Equals(ChromatographicPeak peak)
+        {
+            return SpectraFileInfo.Equals(peak.SpectraFileInfo) 
+                && Identifications.First().ModifiedSequence.Equals(peak.Identifications.First().ModifiedSequence)
+                && ApexRetentionTime == peak.ApexRetentionTime;
         }
 
         public IsotopicEnvelope Apex { get; private set; }
@@ -36,70 +60,17 @@ public ChromatographicPeak(Identification id, bool isMbrPeak, SpectraFileInfo fi
         public int NumIdentificationsByFullSeq { get; private set; }
         public double MassError { get; private set; }
         /// <summary>
-        /// Expected retention time for MBR acceptor peaks (mean)
+        /// Bool that describes whether the retention time of this peak was randomized
+        /// If true, implies that this peak is a decoy peak identified by the MBR algorithm
         /// </summary>
-        public double? RtHypothesis { get; private set; }
-        /// <summary>
-        /// Std. Dev of retention time differences between MBR acceptor file and donor file, used if # calibration points < 6
-        /// </summary>
-        public double? RtStdDev { get; private set;  }
-        /// <summary>
-        /// Interquartile range of retention time differences between MBR acceptor file and donor file, used if # calibration points >= 6
-        /// </summary>
-        public double? RtInterquartileRange { get; private set; }
-
-        public static string TabSeparatedHeader
-        {
-            get
-            {
-                var sb = new StringBuilder();
-                sb.Append("File Name" + "\t");
-                sb.Append("Base Sequence" + "\t");
-                sb.Append("Full Sequence" + "\t");
-                sb.Append("Protein Group" + "\t");
-                sb.Append("Peptide Monoisotopic Mass" + "\t");
-                sb.Append("MS2 Retention Time" + "\t");
-                sb.Append("Precursor Charge" + "\t");
-                sb.Append("Theoretical MZ" + "\t");
-                sb.Append("Peak intensity" + "\t");
-                sb.Append("Peak RT Start" + "\t");
-                sb.Append("Peak RT Apex" + "\t");
-                sb.Append("Peak RT End" + "\t");
-                sb.Append("Peak MZ" + "\t");
-                sb.Append("Peak Charge" + "\t");
-                sb.Append("Num Charge States Observed" + "\t");
-                sb.Append("Peak Detection Type" + "\t");
-                sb.Append("MBR Score" + "\t");
-                sb.Append("PSMs Mapped" + "\t");
-                sb.Append("Base Sequences Mapped" + "\t");
-                sb.Append("Full Sequences Mapped" + "\t");
-                sb.Append("Peak Split Valley RT" + "\t");
-                sb.Append("Peak Apex Mass Error (ppm)");
-                sb.Append("\tMBR Predicted RT");
-                //sb.Append("Timepoints");
-                return sb.ToString();
-            }
-        }
-
-        /// <summary>
-        /// Sets retention time information for a given peak. Used for MBR peaks
-        /// </summary>
-        /// <param name="rtHypothesis"> Expected retention time for peak, based on alignment between a donor and acceptor file </param>
-        /// <param name="rtStdDev"> Standard deviation in the retention time differences between aligned peaks </param>
-        /// <param name="rtInterquartileRange"> Interquartile range og the retention time differences between aligned peaks</param>
-        internal void SetRtWindow(double rtHypothesis, double? rtStdDev, double? rtInterquartileRange)
-        {
-            RtHypothesis = rtHypothesis;
-            RtStdDev = rtStdDev;
-            RtInterquartileRange = rtInterquartileRange;
-        }
+        public bool RandomRt { get; }
+        public bool DecoyPeptide => Identifications.First().IsDecoy;
 
         public void CalculateIntensityForThisFeature(bool integrate)
         {
             if (IsotopicEnvelopes.Any())
             {
-                double maxIntensity = IsotopicEnvelopes.Max(p => p.Intensity);
-                Apex = IsotopicEnvelopes.First(p => p.Intensity == maxIntensity);
+                Apex = IsotopicEnvelopes.MaxBy(p => p.Intensity);
 
                 if (integrate)
                 {
@@ -146,7 +117,7 @@ public void MergeFeatureWith(ChromatographicPeak otherFeature, bool integrate)
                 this.Identifications = this.Identifications
                     .Union(otherFeature.Identifications)
                     .Distinct()
-                    .OrderBy(p => p.PosteriorErrorProbability).ToList();
+                    .ToList();
                 ResolveIdentifications();
                 this.IsotopicEnvelopes.AddRange(otherFeature.IsotopicEnvelopes
                     .Where(p => !thisFeaturesPeaks.Contains(p.IndexedPeak)));
@@ -162,7 +133,40 @@ public void ResolveIdentifications()
             this.NumIdentificationsByBaseSeq = Identifications.Select(v => v.BaseSequence).Distinct().Count();
             this.NumIdentificationsByFullSeq = Identifications.Select(v => v.ModifiedSequence).Distinct().Count();
         }
-
+        public static string TabSeparatedHeader
+        {
+            get
+            {
+                var sb = new StringBuilder();
+                sb.Append("File Name" + "\t");
+                sb.Append("Base Sequence" + "\t");
+                sb.Append("Full Sequence" + "\t");
+                sb.Append("Protein Group" + "\t");
+                sb.Append("Organism" + '\t');
+                sb.Append("Peptide Monoisotopic Mass" + "\t");
+                sb.Append("MS2 Retention Time" + "\t");
+                sb.Append("Precursor Charge" + "\t");
+                sb.Append("Theoretical MZ" + "\t");
+                sb.Append("Peak intensity" + "\t");
+                sb.Append("Peak RT Start" + "\t");
+                sb.Append("Peak RT Apex" + "\t");
+                sb.Append("Peak RT End" + "\t");
+                sb.Append("Peak MZ" + "\t");
+                sb.Append("Peak Charge" + "\t");
+                sb.Append("Num Charge States Observed" + "\t");
+                sb.Append("Peak Detection Type" + "\t");
+                sb.Append("PIP Q-Value" + "\t");
+                sb.Append("PIP PEP" + "\t");
+                sb.Append("PSMs Mapped" + "\t");
+                sb.Append("Base Sequences Mapped" + "\t");
+                sb.Append("Full Sequences Mapped" + "\t");
+                sb.Append("Peak Split Valley RT" + "\t");
+                sb.Append("Peak Apex Mass Error (ppm)" + "\t");
+                sb.Append("Decoy Peptide" + "\t");
+                sb.Append("Random RT");
+                return sb.ToString();
+            }
+        }
         public override string ToString()
         {
             StringBuilder sb = new StringBuilder();
@@ -193,10 +197,12 @@ public override string ToString()
             if (t.Any())
             {
                 sb.Append(string.Join(";", t) + '\t');
+                sb.Append(string.Join(";", Identifications.SelectMany(id => id.ProteinGroups).Select(p => p.Organism).Distinct()) + '\t');
             }
             else
             {
                 sb.Append("" + '\t');
+                sb.Append("" + '\t');
             }
 
             sb.Append("" + Identifications.First().MonoisotopicMass + '\t');
@@ -243,14 +249,16 @@ public override string ToString()
                 sb.Append("" + "MSMS" + "\t");
             }
 
-            sb.Append("" + (IsMbrPeak ? MbrScore.ToString() : "") + "\t");
+            sb.Append("" + (IsMbrPeak ? MbrQValue.ToString() : "") + "\t");
+            sb.Append("" + (IsMbrPeak ? MbrPep.ToString() : "") + "\t");
 
             sb.Append("" + Identifications.Count + "\t");
             sb.Append("" + NumIdentificationsByBaseSeq + "\t");
             sb.Append("" + NumIdentificationsByFullSeq + "\t");
             sb.Append("" + SplitRT + "\t");
             sb.Append("" + MassError);
-            sb.Append("\t" + (IsMbrPeak ? RtHypothesis.ToString() : ""));
+            sb.Append("\t" + DecoyPeptide);
+            sb.Append("\t" + RandomRt);
 
             return sb.ToString();
         }

diff --git a/mzLib/FlashLFQ/FlashLFQ.csproj b/mzLib/FlashLFQ/FlashLFQ.csproj
@@ -1,7 +1,7 @@
 <Project Sdk="Microsoft.NET.Sdk">
 
   <PropertyGroup>
-    <TargetFramework>net6.0</TargetFramework>
+    <TargetFramework>net8.0</TargetFramework>
     <Platforms>x64</Platforms>
   </PropertyGroup>
 
@@ -11,8 +11,11 @@
   </PropertyGroup>
 
   <ItemGroup>
+    <PackageReference Include="CsvHelper" Version="32.0.3" />
     <PackageReference Include="MathNet.Numerics" Version="5.0.0" />
-    <PackageReference Include="NetSerializer" Version="4.1.1" />
+    <PackageReference Include="Microsoft.ML" Version="3.0.1" />
+    <PackageReference Include="Microsoft.ML.FastTree" Version="3.0.1" />
+    <PackageReference Include="NetSerializer" Version="4.1.2" />
     <PackageReference Include="SharpLearning.Optimization" Version="[0.28.0]" />
   </ItemGroup>