Skip to content

Why FERMO?

FERMO is dashboard app for metabolomics data analysis. It aims to make the annotation and prioritization of relevant metabolites convenient, reproducible, and comprehensive.

Supported data types

  • LC-MS(/MS) metabolomics data
  • Genomics data
  • Phenotype (bioactivity) data
  • Environmental/group (meta)data

Supported analyses

  • Molecular feature filtering
  • Spectral similarity (molecular) networking
  • Feature annotation and identification
  • Phenotype data correlation
  • Determination of fold changes
  • Calculation of prioritization scores

Using FERMO in data analysis

In a typical metabolomics analysis, hundreds or thousands of molecular features are routinely detected. This wealth of information makes it challenging to manually identify and prioritize relevant molecules for further investigations. FERMO supports the prioritization task by an automated annotation pipeline, calculation of prioritization scores, and an intuitive graphical user interface. This allows to rapidly inspect data, formulate hypotheses, and determine relevant features.

Benefits of FERMO

  • Interface: FERMO comes as intuitive dashboard allowing to prioritize data with a few clicks.
  • Flexibility: FERMO only requires a feature peaktable - everything else is optional.
  • Tool integration: FERMO integrates many different tools into a single pipeline.
  • Modularity: FERMO is fully modular and allows for easy enabling/disabling of functionality.
  • Reproducibility: FERMO's input parameter file allows for reproducible analysis runs.
  • Prioritization scores: FERMO combines the individual modules in meta-scores useful for prioritization.

Limitations of FERMO

  • Pre-processing: FERMO does not accept raw data and needs a pre-processed feature peaktable.
  • Data compatibility: FERMO needs DDA-ESI-LC-MS(/MS) input data. Currently, accepted input file formats are limited.
  • Scale: FERMO's interface is not intended to display a large number (>100) of samples. While there is not fixed threshold, users intending to run large-scale analyses are advised to look into running fermo_core locally.