release 1.0.6

• major update to the differential detection metric
  • computation of z-scores was further refined
  • additional parameters via analysis_quickstart() give more control
  • proteins with just 1 peptide are now skipped by default
  • The analysis_quickstart() example on the main GitHub page has been updated accordingly
  • Additional details can be found in the differential detection vignette
• new function for summarizing statistical results at gene-level; summarise_stats() can merge results from DEA and differential detection, returning 1 result/statistic per gene (across contrasts and DEA algorithms).
  • Additional details can be found in the summarizing results from statistics + gene ID mapping vignette
• See the "summarizing results from statistics + gene ID mapping" vignette for more details
• new function for mapping mouse, rat and human proteins to human proteins, e.g. to prepare your data for downstream analyses where human gene identifiers are required. This requires external/downloaded mapping tables.
  • Additional details can be found in the summarizing results from statistics + gene ID mapping vignette
• new function for merging technical replicates up-front; merge_replicate_samples() . Usage is situational, in some cases you may want to deal with (technical) replicates in downstream statistical modeling. Example and documentation @ ?merge_replicate_samples()
• the MS-DAP dataset object is now stored in the output directory by default ('dataset.RData')

release 1.0.5

• reworked import function for ProteomeDiscoverer (after installing this MS-DAP update, see the function documentation for import_dataset_proteomediscoverer_txt for details and options)
• bugfix for importing Spectronaut reports that use multiple spectral libraries
• various minor updates

release 1.0.4

This release brings a major update to our support for FragPipe datasets, if you're using FragPipe make sure to upgrade MS-DAP:

• 'recent' FragPipe versions produce different output files, we have updated MS-DAP accordingly
• MS-DAP now supports 3 different FragPipe LFQ dataset output formats
• See further this documentation on using FragPipe and MS-DAP together
• To update your MS-DAP installation, refer to this brief instruction on the MS-DAP main page

release 1.0.3.1

minor update to resolve issues with DIA-NN report.tsv files that contain multiple rows of data for a precursor ID in the same sample, but with zero values in the retention time and quantity/intensity columns (which would result in a MS-DAP error message peptide_id*sample_id combinations are not unique ... in version 1.0.3)

release 1.0.3

• breaking change; some parameters have been renamed
  mostly applies to users writing code that directly invokes dea/normalization functions, or operates on the dataset$de_proteins table
  • renamed algo_de to dea_algorithm throughout MS-DAP codebase
  • renamed dea_protein_rollup and algo_rollup to rollup_algorithm throughout MS-DAP codebase
  • refactored parameter names for normalize_matrix()
• normalization; a new variant of VWMB is now available, Mode Within Mode Between (MWMB). Normalize (/scale) samples within each sample group such that their pairwise log-foldchange modes are zero (whereas VWMB would normalize replicates by reducing peptide variation), then scales between groups such that the log-foldchange mode is zero (i.e. the between-group part is the same as VWMB). If the dataset has (unknown) covariates and a sufficient number of replicates, this might be beneficial because covariate-specific effects are not averaged out as they might be with VWMB. However, our recommendation for general-purpose (i.e. good start on initial investigation of a dataset) is still norm_algorithm = c("vsn", "modebetween_protein")
• peptide-to-protein rollup; Tukey's median polish (TMP) is now also available in analysis_quickstart() and dea() through the rollup_algorithm parameter. MaxLFQ remains default; our benchmarking showed its marginally better overall than TMP
• report; the foldchange-distribution visualizations of within-group outliers are now accompanied by a table with respective scores
• report; configured normalization and rollup algorithms are now used in most QC figures in the PDF report (instead of always using VWMB for the retention-time-error and CoV plots)
• export data; "peptide*sample" data matrices are now also exported when using analysis_quickstart() with parameter output_abundance_tables=TRUE. If you're not using the analysis_quickstart() function, you may directly call export_peptide_abundance_matrix() to create "peptide*sample" TSV files
• docker; the MS-DAP docker container is now built on recent R version 4.2.1 (and added minor improvements to the unix launcher script)
• MS-DAP can now import compressed fasta files and data tables while importing from MaxQuant / DIA-NN / Spectronaut / FragPipe / MetaMorpheus. Supported compression formats; .zip|.gz|.bz2|.xz|.7z|.zst|.lz4
  • you can present compressed files as input, or provide file paths as-is without the extension for the archive (e.g. filename="C:/experiment_x/diann/report.tsv"). If the input file is not found, MS-DAP will automatically check if a compressed variant of the file is available (by matching against supported file extensions)
  • for upstream software that produces a folder with multiple files that need to read into MS-DAP; compress each file individually (e.g. for MaxQuant, make separate zip/gzip/Zstandard/etc. files like evidence.txt.zip, proteinGroups.txt.zip and peptides.txt.zip , then work with MS-DAP as usual)
• you can now skip the creation of any output files/directories in analysis_quickstart() by setting output_dir=NA
• more documentation was added to both R functions and the GitHub vignettes
• bugfix for the DEqMS package, resolves a rare error when dealing with proteins that have (near) zero variation
• various few minor bugfixes and code speedups (no impact on analysis results)

release 1.0.2

• new implementation for bootstrapped inference of foldchange thresholds in DEA analysis, which scales to large datasets
  • this applies to users who set dea_log2foldchange_threshold=NA in analysis_quickstart() to let an algorithm infer a threshold/cutoff for log2 foldchanges in volcano plots
  • importantly, be aware that this may result in a minor change in the estimated foldchange threshold (0.1~3% difference in the estimated threshold for 10 datasets we tested)
  • keep an eye on this when comparing analyses across MS-DAP versions
  • on that note; in datasets with few replicates (for instance WT~KO with 3 replicates each), a bootstrap analysis to find background-level foldchanges based on permutations of sample-to-group assigments is limited and possibly over-conservative (i.e. there aren't many permutations to make). It's recommended to apply this approach to datasets with 5+ replicates
• additional input validations for volcano plots (for users that call this function directly, i.e. outside of canonical analysis_quickstart() workflow)
• expanded QC plots for within-group variation, peptide foldchange distribution plots now come in 2 styles and have improved color-coding
• bugfix specific to R 4.x (a warning that was raised to an error only in R 4.x)
• MS-DAP has been tested against the new R release 4.2
• documentation updates

release 1.0.1

minor update:

• updated EncyclopeDIA import function #1
• updated user-guide for importing Spectronaut datasets #2
• new msdap::plot_volcano() function to create volcano plots with custom labels post-analysis #3
• new variance-explained analysis (experimental feature atm), optionally accessible through msdap::analysis_quickstart()
• a few minor bugfixes (cosmetics / user-convenience only, no impact on analysis results)

release 1.0

first production release

beta 0.2.8.2

minor bugfixes

beta 0.2.8.1

• reworked FragPipe import functions, now supports IonQuant (make sure to assign Experiment IDs in the workflow tab of FragPipe)
• protein rollup by MaxLFQ, implementation from iq package is now default
• reworked utility function to export protein abundance matrices
• various minor updates to data visualizations and documentation