Metabolomics Navigator Tools, strategies & workflow guidance for metabolomics

Quality control and filtering

pyEVA

A Python implementation of EVA for automated evaluation of extracted-ion chromatogram fidelity, using deep learning to classify LC-MS features as likely true or false signals.

LC-MSUntargeted metabolomicsCommand lineOpen source
Access toolSuggest an update

Overview

pyEVA is a Python implementation of the EVA feature-fidelity workflow. It extracts chromatographic traces around supplied LC-MS features, creates EIC plots and classifies them to assist detection of false or unreliable features.

Also known as: EVA

What can this tool do?

Feature selection

Common uses

  • Automatically classify extracted-ion chromatograms as higher- or lower-fidelity metabolomic features.
  • Filter feature tables after chromatographic peak detection.

Scientific fit

Primary function

Quality control and filtering

Other functions

Machine learning and prediction

Analysis types

Untargeted metabolomics

MS level

MS1

Acquisition

Full-scan MS

Ion mobility

Not specified

Data compatibility

Inputs

Raw instrument data, Feature or peak table

Outputs

Feature or peak table, Human-readable report

Input formats

mzML, mzXML, CSV or TSV

Output formats

CSV or TSV

Scope and considerations

Designed for

  • Automated feature-fidelity assessment of LC-MS extracted-ion chromatograms.

Not designed for

  • Molecular identification or a replacement for acquisition-aware review of low-abundance and unusual chromatographic peaks.

Important requirements

  • The documented Python environment plus R/xcms dependencies, raw mzML or mzXML files and a CSV whose first two columns are m/z and retention time.

Considerations

  • Smoothing and peak-shape domain affect classifications. Inspect EICs and prediction outcomes rather than automatically interpreting every rejected feature as an artifact.

Access and documentation

Primary accessDocumentation

MIT source distribution. The cited EVA paper describes the original method; the repository documents the later Python implementation and its setup.

License: MIT

Evidence and review

Enveda intake rechecked against official README/documentation, repository access terms and primary publication metadata on 2026-10-27. Editorial source review only; no software execution or independent benchmarking. See docs/enveda-expansion-review-2026-10-27.md.

Core publications

Resource papers, direct updates and tool-specific methods. Selection policy.

  1. EVA: Evaluation of Metabolic Feature Fidelity Using a Deep Learning Model Trained With Over 25000 Extracted Ion Chromatograms. Analytical Chemistry (2021). doi:10.1021/acs.analchem.1c01309

    Original EVA method paper; the official pyEVA repository documents its Python reimplementation.

Related tools

dbnormSimilar function · Compare statistical batch-correction models and inspect technical variation in metabolomics matrices.InjectionDesignSimilar function · A web service for designing LC-MS and GC-MS multi-omics injection sequences with QC placement, inter-batch balancing, intra-batch randomization, and downloadable worksheets.IonToolPackSimilar function · IonToolPack groups graphical mass-spectrometry utilities for raw-data visualization, PeakQC quality assessment, tandem-library matching, targeted MS1 extraction and feature-list comparison.ISFragSimilar function · ISFrag recognizes candidate in-source fragments in LC-MS data by combining feature relationships and available MS/MS evidence.

Entry history

Changes to this catalogue entry, not the tool's software release history. Acceptance and editorial review are separate.

  1. 2026-10-27 · First committed entry

    Add 27 source-reviewed tools from Enveda intake

    View change · Entry at this revision

Full GitHub history