tpptrSplineFitAndTest: Perform spline fitting and analyze by moderated F-test
In TPP: Analyze thermal proteome profiling (TPP) experiments

Description Usage Arguments Details Value See Also Examples

A wrapper function around the functions tpptrFitSplines, tpptrFTest, tpptrPlotSplines, which fits natural splines to all proteins in a dataset and detect differential behavior between conditions by a moderated F-test. The results are formatted as a wide table with one row per protein. This table contains all the original data, the test results, and (optionally) additional annotation columns for each protein.

tpptrSplineFitAndTest(
  data,
  factorsH1,
  factorsH0 = character(),
  resultPath = NULL,
  doPlot = TRUE,
  nCores = "max",
  splineDF = 3:7,
  additionalCols = NULL,
  verbose = NULL,
  ggplotTheme = NULL
)

`data`	the data to be fitted.
`factorsH1`	which factors should be included in the alternative model?
`factorsH0`	which factors should be included in the null model?
`resultPath`	location where to store the spline plots per protein.
`doPlot`	boolean value indicating whether melting curves should be plotted, or whether just the curve parameters should be returned.
`nCores`	either a numerical value given the desired number of CPUs, or 'max' to automatically assign the maximum possible number (default).
`splineDF`	degrees of freedom for natural spline fitting.
`additionalCols`	additional annotation per protein to append to the result table.
`verbose`	DEPRECATED
`ggplotTheme`	DEPRECATED.

Plots of the natural spline fits will be stored in a subfolder with name Spline_Fits at the location specified by resultPath.

Argument data can either be long table, or a list of expressionSets as returned by tpptrImport. If a long table, it needs to contain the following columns: 'uniqueID' (identifier), 'x' (independent variable for fitting, usually the temperature) and 'y' (dependent variable for fitting, usually the relative concentration).

Argument splineDF specifies the degrees of freedom for natural spline fitting. As a single numeric value, it is directly passed on to the splineDF argument of splines::ns. Experience shows that splineDF = 4 yields good results for TPP data sets with 10 temperature points. It is also possible to provide a numeric vector. In this case, splines are fitted for each entry and the optimal value is chosen per protein using Akaike's Information criterion.

A data frame in wide format with one row per protein. It contains the smoothing spline parameters and F-test results obtained by comparing the null and alternative models.

ns, AICc

data(hdacTR_smallExample)
tpptrData <- tpptrImport(configTable = hdacTR_config, data = hdacTR_data)
fitData <- tpptrTidyUpESets(tpptrData)
hdacSplineFits <- tpptrSplineFitAndTest(data = fitData,
                                        factorsH1 = "condition",
                                        nCores = 1,
                                        splineDF = 4:5,
                                        doPlot = FALSE)
# Show estimated splines for HDAC1:
filter(hdacSplineFits, Protein_ID == "HDAC1")
# -> Which proteins showed significant condition effects?
hdacSplineFits %>% filter(p_adj_NPARC <= 0.01) %>% select(Protein_ID, p_adj_NPARC)
# Quality control: test for replicate-specific effects:
 testResults <- tpptrSplineFitAndTest(data = fitData,
                                     factorsH1 = "replicate",
                                     nCores = 1,
                                     splineDF = 4,
                                     doPlot = FALSE)
# -> Which proteins showed significant replicate effects?
testResults %>% filter(p_adj_NPARC <= 0.01) %>% select(Protein_ID, p_adj_NPARC)