← All apps

TPP Solver

Thermal Proteome Profiling analysis. Drop a FragPipe intensity table (TSV) and a metadata file (CSV) to fit sigmoidal melting curves, extract each protein's melting temperature (T_m), and test for stability shifts between treatments. Everything runs in your browser — your data never leaves your machine.

Intensity table (TSV)

FragPipe output — a Protein ID column plus one intensity column per sample. Drop or click to choose.

Metadata (CSV)

Columns Temperature, Treatment, Samples. Drop or click to choose.

Analysis setup

Normalize to a reference temperature

Divides each curve by its intensity at the reference temperature, so every curve starts near 1.0 (fully soluble).

Reference temperature (°C)

R² threshold: 0.80

Minimum R² for a curve fit to be accepted.

Random seed (imputation)

Missing/zero intensities are imputed with small random values. Fixing the seed makes results reproducible.

Statistics & visualizations

Treatment 1 (control)

Treatment 2

Mann-Whitney U + Benjamini-Hochberg Per-protein test of T_m shift between the two treatments, with FDR correction and a volcano plot.

Distribution of melting points Histogram of all fitted T_m values.

Violin plot by treatment Compare the T_m distributions of each treatment.

Summary

Melting curve

Volcano plot — melting-point shifts

significant (FDR < 0.05) not significant FDR = 0.05

Statistical results

Distribution of melting points

Melting points by treatment

Averaged summary

Melting curves are fit with the three-parameter sigmoid f(T) = a / (1 + e^-(T-b)) + c by Levenberg-Marquardt least squares; the melting temperature is the inflection point b. Fits below the R² threshold are discarded. Significance uses a two-sided Mann-Whitney U test on per-replicate T_m values (exact for small samples, normal approximation with tie and continuity correction otherwise), with Benjamini-Hochberg FDR control. This is a browser port of tppsolver.gnps2.org; the Gene Ontology annotation and Shapiro-Wilk panels of the original are omitted because they require a server-side database.