Proteomics > Which service should I request? > Proteome Identification and Quantification > Proteome Quantification - Label-Free

Proteome Quantification - Label-Free (LFQ)

General description

Label-free proteomics quantification (LFQ) is an established approach to relatively quantify proteins on a large dataset in a rapid, reproducible, flexible, and affordable manner. Examples of applications are the identification of differentially regulated proteins and pathways in conditions such as:

• Treatment(s) vs Control
• Disease vs Healthy
• KO vs WT
• Time series
• ...and more

Depending on the experimental design and the biological question, we also offer Label-based proteome quantification methods

Workflow

Our default workflow for label-free proteome quantification consists of 3 steps.

A. Planning

As a first step, we need to clarify the goal of the project and clarify the analytic and bioinformatics strategy. A proper experimental design will be defined together with FGCZ staff. Please contact us at proteomics at fgcz.ethz.ch for a meeting, or directly contact your coach If you already have a B-Fabric project at FGCZ.

B. Pilot (QC) step

Goal:

• Test the feasibility of the experimental procedure (protocol, starting amount, quality of the samples, ...)
• estimate the sample's size

Depending on the type of your sample (e.g., pull-down, supernatant, whole-cell lysate) we observe some dozens up to a few thousands of proteins. While the overall number of proteins can highly vary depending on the type of experiment, it is crucial that the number of proteins between your biological replicates is similar (reproducibility). This pilot step helps us determining this number of replicates necessary to detect the protein fold-change (effect size) that you would like to detect between two conditions, e.g. a reference and a treatment. In general, Smaller biologically significant effect sizes require more samples to obtain a statistically significant result.

Four samples, corresponding to four biological replicates of your sample of interest, will be processed and measured in parallel. This is step is highly recommended!

C. Label Free Quantification step (LFQ)

The main LFQ experiment is performed following the protocol tested and optimized during the QC step. A minimum of 8 samples (4 samples, 2 categories) are processed and the protein quantification is
performed. More biological replicates and more categories could be included in the experiment depending on the results of the sample size estimation.
Dozens of samples can be analyzed in as single experiment (up to 100s).

Requirements and considerations

• The first QC step is mandatory. Do not prepare the samples for the main LFQ until these results are available.
• Ideal protein amount: 25-50 µg (minimum 1 µg)
• Replicates: 4+ for in-vitro experiments, 5/10+ for animal/human experiments (depending on the results of the sample size estimation)
• Buffer composition: every experiment requires some optimization, please contact us at proteomics at fgcz.ethz.ch
• Turnaround time
  • QC (Pilot) step: ~1 week
  • Main label-free quantification experiment (2-4 weeks)

Data acquisition and data analysis

The data can be acquired and analysed in multiple ways, but these are the 2 most common strategies used at FGCZ:

1. Data-Dependent Acquisition (DDA) mode

Valid for the majority of the LFQ experiments at FGCZ:

• data acquisition on M-Class + Orbitrap Fusion Lumos (or similar generation, see here )
• DDA mode, with randomized data acquisition
• iRT spiked in the samples (Biognosys - QC purposes)
• QC sample every 4-5 runs
• Protein identification and quantification using MaxQuant (quantification at the MS1 level)
• Data analysis using linear or mixed statistical models
• Generation of reports, tables and plots
• (for selected species, e.g. human and mouse, Gene Set Enrichment Analysis (GSEA) and Over-Representation Analysis (ORA) are performed using our fgcz.gsea.ora package, which integrates WebGestalt)

2. Data-Independent Acquisition (DIA) mode

Mainly used for large datasets, clinical samples or for peptide-centric quantification

• data acquisition on M-Class + Orbitrap Fusion Lumos (or similar generation, see here
• DIA mode, with randomized data acquisition (incl. 4-5 samples in DDA mode)
• iRT spiked in the samples
• QC sample every 4-5 runs
• Generation of mixed libraries (from DIA and DDA data) using Spectronaut (Biognosys)
• Protein identification and quantification using Spectronaut (quantification at the MS2 level)
• Data analysis using linear or mixed statistical models
• Generation of reports, tables and plots
• (for selected species, e.g. human and mouse, Gene Set Enrichment Analysis (GESA) and Over-Representation Analysis (ORA) are performed using our fgcz.gsea.ora package, which integrates WebGestalt. See FGCZ-GSEA-ORA Poster

Assessment of the LC-MS performances

The performances of the LC-MS instruments are constantly monitored using 3 different pipelines (PanoramaWeb; QCloud; internal FGCZ tool(s)), and 2 sets of standards:

• autoQC4L:
  • 25 ng of digested proteins from K652 cells + 6x5 LCMSMS Peptide Reference Mix (Promega)
  • at the start and end of a queue, to assess identification and quantification rates (see QC4L poster)
• autoQC01:
  • 10 peptides from commercial digested BSA (25 fmol) and 11 iRT peptides
  • PRM acquisition - acquired every 4-6 samples to monitor potential drops in sensitivity.