In lgatto/synapter: Label-free data analysis pipeline for optimal identification and quantitation

knitr::read_chunk(system.file("reports", "scripts", "synergise.R",
                              package="synapter"))

Input {.tabset}

Files

inputFiles <-
    rbind(c("Identification peptide:",   basename(obj$IdentPeptideFile),  nrow(obj$IdentPeptideData)),
          c("Identification fragments:", basename(obj$IdentFragmentFile), length(obj$IdentFragmentData)),
          c("Quantitation peptide:",     basename(obj$QuantPeptideFile),  nrow(obj$QuantPeptideData)),
          c("Quantitation Pep3D:",       basename(obj$QuantPep3DFile),    nrow(obj$QuantPep3DData)),
          c("Quantitation Spectra:",     basename(obj$QuantSpectrumFile), length(obj$QuantSpectrumData)),
          c("Fasta file:",    basename(obj$DbFastaFile), ""))
colnames(inputFiles) <- c("", "File name", "Number of features")
knitr::kable(inputFiles, align="llr")

General Parameters

parameters <-
    rbind(c("Master", master),
          c("Peptide FDR", fdr),
          c("Protein FPR", fpr),
          c("Peptide Length", peplen),
          c("Missed cleavages", missedCleavages),
          c("Isoleucin equals Leucin", IisL),
          c("Identification mass tolerance (ppm)", identppm),
          c("Quantitation mass tolerance (ppm)", quantppm),
          c("Loess span (RT modelling)", span.rt),
          c("Loess span (Intensity modelling)", span.int),
          c("Filtering unique peptides", as.character(uniquepep)))
colnames(parameters) <- c("Parameter name", "Value")
knitr::kable(parameters, align="lr")

Grid Search Parameters

gridParameters <-
    rbind(c("Mass tolerance (ppm) start", grid.ppm.from),
          c("Mass tolerance (ppm) end", grid.ppm.to),
          c("Mass tolerance (ppm) by", grid.ppm.by),
          c("Retention time stdev start", grid.nsd.from),
          c("Retention time stdev end", grid.nsd.to),
          c("Retention time stdev by", grid.nsd.by),
          c("Ion mobility start", grid.imdiffs.from),
          c("Ion mobility end", grid.imdiffs.to),
          c("Ion mobility by", grid.imdiffs.by),
          c("Feature proportion ", grid.subset),
          c("Number of features", grid.n),
          c("parameters selection", grid.param.sel))
colnames(gridParameters) <- c("Grid parameter", "Value")
knitr::kable(gridParameters, align="lr")

Fragment Matching Parameters

fmParameters <-
    rbind(c("Fragment Matching mass tolerance (ppm)", fm.ppm),
          c("Minimal Intensity for Identification Fragments", fm.ident.minIntensity),
          c("Minimal Intensity for Quantitation Spectra", fm.quant.minIntensity),
          c("Minimum Number of Common Peaks", fm.minCommon),
          c("Minimum Number of Delta Common Peaks", fm.minDelta),
          c("Unique Matches FDR", fm.fdr.unique),
          c("Non-Unique Matches FDR", fm.fdr.nonunique))
colnames(fmParameters) <- c("Parameter name", "Value")
knitr::kable(fmParameters, align="lr")

Filtering

Peptide False Discovery Rate

Peptide Scores

Peptide Numbers

Peptide FDR

Mass tolerance

Retention Time Modelling

Retention Time Model

Feature Space

plotFeatures(obj, what="all", ionmobility=TRUE)

Grid Optimisation

r if(grid.n) { paste0("Using", grid.n, "features.") } else { paste0("Using ", grid.subset * 100, "% of the data.") }

# abuse inline r here to show the progress bar of the gridsearch in the
# R console - regular r chunks capture stdout

r if (grid.n) { searchGrid(obj, ppms=seq(grid.ppm.from, grid.ppm.to, grid.ppm.by), nsds=seq(grid.nsd.from, grid.nsd.to, grid.nsd.by), imdiffs=seq(grid.imdiffs.from, grid.imdiffs.to, grid.imdiffs.by), n=grid.n, verbose=verbose) } else { searchGrid(obj, ppms=seq(grid.ppm.from, grid.ppm.to, grid.ppm.by), nsds=seq(grid.nsd.from, grid.nsd.to, grid.nsd.by), imdiffs=seq(grid.imdiffs.from, grid.imdiffs.to, grid.imdiffs.by), subset=grid.subset, verbose=verbose) }

Grid Tables {.tabset}

names <- c("total", "model", "details")
what <- c("total", "model", "details")
for (i in seq(along=names)){
  cat("\n####", names[i], "\n")
  cat("\n##### Image\n")
  plotGrid(obj, what=what[i])
  cat("\n\n##### Tables {.tabset}\n")
  for (j in 1:dim(getGrid(obj)[[i]])[3]) {
    cat("\n###### imdiff:", dimnames(getGrid(obj)[[i]])[[3]][j], "\n")
    print(knitr::kable(as.data.frame(t(getGrid(obj)[[i]][,,j]))))
  }
  cat("\n##### Best Grid Parameters\n")
  cat("\nBest Grid Value ", names[i], ": ", getBestGridValue(obj)[i], "\n", sep="")
  print(knitr::kable(as.data.frame(getBestGridParams(obj)[[i]])))
}

Grid Details

gridDetails <- do.call(rbind, getGridDetails(obj))
cond <- do.call(rbind, strsplit(rownames(gridDetails), ":"))
colnames(cond) <- c("nsd", "ppm", "imdiff")
gridDetails <- cbind(cond, gridDetails)
gridDetails[] <- as.numeric(gridDetails)
rownames(gridDetails) <- NULL
knitr::kable(as.data.frame(gridDetails))

Best Grid Parameters

Setting best grid parameters using 'r grid.param.sel'

• Mass tolerance (ppm): r obj$PpmError
• Number of retention time stdev: r obj$RtNsd

EMRT Matching

findEMRTs(obj)
plotEMRTtable(obj)

Fragment Matching

# abuse inline r here to show the progress bar of the fragment matching in the
# R console - regular r chunks capture stdout

Filtering Fragments

plotCumulativeNumberOfFragments(obj, what = "fragments.ident")
plotCumulativeNumberOfFragments(obj, what = "spectra.quant")

r invisible(filterFragments(obj, what="fragments.ident", minIntensity=fm.ident.minIntensity)) r invisible(filterFragments(obj, what="spectra.quant", minIntensity=fm.quant.minIntensity))

r invisible(fragmentMatching(obj, ppm = fm.ppm)) r invisible({pdf(file.path(outputdir, "FragmentMatching.pdf"), width=10, height=8); plotFragmentMatching(obj); dev.off()})

Performance {.tabset}

fragmentMatchingStats <- plotFragmentMatchingPerformance(obj)

names <- c("Unique Matches", "Non-Unique Matches")
firstColumn <- c("Number of Common Peaks", "Delta Common Peaks")
for (i in seq(along=names)){
  cat("\n####", names[i], "\n")
  tab <- as.data.frame(fragmentMatchingStats[[i]])
  colnames(tab) <- c(firstColumn[i], toupper(colnames(tab)[-1]))
  print(knitr::kable(tab))
}

Best Parameters

common <- c(NA, NA)
selCommon <- which(fragmentMatchingStats$unique[, "ncommon"] >= fm.minCommon &
  fragmentMatchingStats$unique[, "fdr"] <= fm.fdr.unique)[1]

if (!is.na(selCommon)) {
  common <- fragmentMatchingStats$unique[selCommon, c("ncommon", "fdr")]
  filterUniqueMatches(obj, minNumber = common[1L])
}

delta <- c(NA, NA)
selDelta <- which(fragmentMatchingStats$nonunique[, "deltacommon"] >= fm.minDelta &
  fragmentMatchingStats$nonunique[, "fdr"] <= fm.fdr.nonunique)[1]

if (!is.na(selDelta)) {
  delta <- fragmentMatchingStats$nonunique[selDelta, c("deltacommon", "fdr")]
  filterNonUniqueMatches(obj, minDelta = delta[1L])
}

filterNonUniqueIdentMatches(obj)

tab <- cbind(paste(names, firstColumn, sep=" - "), rbind(common, delta))
colnames(tab) <- c("Type", "Value", "FDR")
rownames(tab) <- NULL
print(knitr::kable(tab))

EMRT Rescue

rescueEMRTs(obj, mergedEMRTs)
plotEMRTtable(obj)

## @knitr synergise.emrt.table
tab <- as.data.frame(getEMRTtable(obj))
colnames(tab) <- c("Number of assigned EMRTs", "Freq")
knitr::kable(tab)

Intensity Model

plotIntensity(obj, what="data")

setLowessSpan(obj, span.int)
modelIntensity(obj)
plotIntensity(obj, what="model", nsd=1) ## better focus on model

Performance

Enrichment

Overlap

Source

Exported Files

• Identification Peptides
• Quantitation Peptides
• Merged Peptides
• Matched Peptides
• QuantPeptides.csv
• Grid Details
• Fragment Matching Plots
• Saved R Object (binary)

Full Analysis Log

Session Info

lgatto/synapter documentation built on Sept. 28, 2022, 6:53 a.m.