UC Davis Proteomics Core Facility
Upload your DIA-NN proteomics data and instantly identify differentially expressed proteins with interactive volcano plots, pathway enrichment, and AI-powered interpretation. Supports phosphoproteomics, multi-omics integration, and core facility management.
Video tutorials, educational resources, training courses, and the latest news in proteomics research.
Data analysis walkthroughs, DIA-NN tutorials, Spectronaut guides, and proteomics short course lectures from the UC Davis Proteomics Core Facility.
Browse all videosState-of-the-art mass spectrometry services including label-free quantitative profiling, DIA, post-translational modification analysis, amino acid analysis, and Edman sequencing.
proteomics.ucdavis.eduStay current with the latest developments, publications, and breakthroughs in proteomics research. Curated news and commentary from the field.
proteomicsnews.blogspot.comIntensive summer training covering mass spectrometry fundamentals, data acquisition, and analysis. Learn DIA-NN, Spectronaut, MaxQuant, and more from expert instructors.
View course detailsExplore the key citations behind DE-LIMP's methodology. Use Google's NotebookLM to dive deep into the limpa, limma, and DIA-NN papers.
Explore methodology citationsIf your core facility already processed your samples with DIA-NN, you just need the report.parquet file from their output. Upload it to DE-LIMP, assign your experimental groups, and the app handles normalization, statistics, and visualization. No programming required.
A complete proteomics analysis platform — from raw data search to multi-omics integration and core facility management.
Compare two analyses of the same dataset across tools (DE-LIMP vs DE-LIMP, Spectronaut, or FragPipe). Four diagnostic views highlight differences between your analyses, with plain-English explanations for why specific proteins disagree. Optionally use multi-omics factor analysis (MOFA2) or AI to explore patterns in the disagreements.
Check data quality before running time-consuming searches. The app reads total ion current (TIC) profiles from your timsTOF raw files and automatically flags problems like failed injections, sample carryover between runs, retention time shifts, and uneven sample loading. Three views: faceted, overlay, and metrics.
Full audit trail for every DIA-NN search (26 parameters) and pipeline run. Import Settings or Results from past searches. Organize analyses into projects with summary cards and cross-reference navigation.
Automatic contaminant detection with per-sample breakdown, keratin flagging, intensity heatmap, and Expression Grid highlighting. Monitor contamination levels across your experiments.
Explore data without DE analysis. Quartile-based abundance profiles reveal proteins that shift rank across samples. Sample-sample scatter shows pairwise correlations with outlier labeling. Works with no-replicates mode.
Visual directory browser for remote HPC files via SSH. Click to navigate, color-coded file types, breadcrumb navigation. Load completed search results directly from the cluster.
Download protein FASTA databases from NCBI for non-model organisms. Automatic gene symbol mapping via E-utilities batch lookup. Complements UniProt for species with better NCBI coverage.
Run DIA-NN searches via Docker (local) or HPC (SSH/SLURM) with parallel 5-step pipelines, spectral library caching, FASTA database library, automatic queue management (detects when you've hit your CPU allocation and switches partitions), and FASTA database library.
Manage your proteomics lab from one dashboard. Track search jobs, monitor instrument performance over time, auto-fill staff credentials, and generate standalone HTML reports for clients.
Find hidden patterns across multiple data types. Combine 2-6 datasets (e.g., proteomics + phospho + RNA-seq) to discover which biological signals are shared and which are unique to each measurement.
Site-level differential expression, KSEA kinase activity inference, and motif analysis for phospho-enriched data. Auto-detection of phosphosites from DIA-NN output.
Fully interactive Plotly volcano plots with correct P.Value/adj.P.Val handling, DE protein count annotations, and click or box-select to highlight across all views.
Ask AI questions about your proteomics results. Get automatic experiment summaries, have interactive conversations about your data, and export for Claude or ChatGPT. Powered by Google Gemini (free API key required).
Rank-based GSEA across Gene Ontology categories (Biological Process, Molecular Function, Cellular Component) and KEGG pathways, with dot plots, enrichment maps, ridgeplots, and automatic organism detection for 12 species.
Download your complete analysis as a portable .zip package for deep analysis with AI assistants. Standalone HTML reports, one-click CSV exports, reproducibility code logs, and full session save/load.
Tabbed dashboard with Volcano + Heatmap, Results Table, PCA, and interactive CV Analysis scatter plots (logFC vs Avg CV) all in one view.
Up and running in under five minutes.
Run natively with R 4.5+. Auto-installs all packages on first launch.
One command: docker compose up. Includes DIA-NN search engine.
Try instantly at huggingface.co. No install needed.
DE-LIMP requires R version 4.5 or newer for Bioconductor 3.22+ compatibility. Download from cloud.r-project.org.
git clone https://github.com/bsphinney/DE-LIMP.git
cd DE-LIMPOr download as a ZIP from the GitHub page (click the green "Code" button, then "Download ZIP") and extract it.
The app will automatically install any missing packages on first launch.
Rscript -e "shiny::runApp('.', port = 3838, launch.browser = TRUE)"Use the sidebar to upload a DIA-NN report.parquet file, assign experimental groups, and run the analysis pipeline. Or click "Load Example Data" to try with a built-in dataset.
DIA-NN is free for academic use but cannot be redistributed. Build it locally:
# Mac/Linux:
bash build_diann_docker.sh
# Windows PowerShell:
.\build_diann_docker.ps1docker compose upOpen http://localhost:3838. Place raw files in data/raw/ and FASTA in data/fasta/.
| Requirement | Version | Notes |
|---|---|---|
| R | 4.5+ | Required for Bioconductor 3.22+ |
| Bioconductor | 3.22+ | Automatically configured with R 4.5 |
| limpa | Latest | Core analysis engine (auto-installed) |
| limma | Latest | Statistical framework (auto-installed) |
| MOFA2 | Latest | Multi-omics factor analysis (auto-installed) |
| Shiny + bslib | Latest | Web framework (auto-installed) |
| Docker | 20+ | Optional — for DIA-NN search integration |
DE-LIMP uses the LIMPA package to run a well-established proteomics analysis workflow:
Upload a DIA-NN report.parquet file, or search raw files directly with DIA-NN via the integrated search tab. Supports spectral library caching and custom FASTA sequences.
Correct for technical variation between mass spec runs (e.g., differences in sample loading or instrument performance) using DPC-CN normalization, so that observed intensity differences reflect biology, not instrumentation.
The mass spectrometer measures individual peptide fragments (precursors). This step combines those measurements into a single abundance value for each protein using DPC-Quant (Detection Probability Curve Quantification), which models missing values probabilistically rather than imputing them, accurately estimating protein quantities from peptide data.
Compare protein levels between your experimental groups (e.g., treated vs. control) using limma, a statistical method that borrows information across all proteins to produce reliable p-values even with few replicates. Results are corrected for multiple testing (FDR) so you can be confident the hits are real.
Site-level differential expression for phospho-enriched datasets. Auto-detects phosphosites from DIA-NN reports or accepts pre-computed site matrices.
Gene set enrichment analysis (GO, KEGG) to identify affected pathways. Multi-omics factor analysis to integrate proteomics, phosphoproteomics, transcriptomics, and more.