Published June 29, 2017 | Version v1
Dataset Open

Data used in "Fast metabolite identification with Input Output Kernel Regression"

  • 1. Department of Computer Science, Aalto University & Helsinki Institute for Information Technology, Espoo, Finland
  • 2. Chair for Bioinformatics, Friedrich-Schiller University, Jena, Germany
  • 3. LTCI, CNRS, Télécom ParisTech, Université Paris-Saclay, Paris, France


This repository contains the data used in [1] to evaluate the performance for metabolite identification from tandem mass spectra. These data have been extracted and processed in [2]. We used a subset of 4138 MS/MS spectra extracted from the GNPS public spectral library ( for training and evaluation. For searching, we used molecular structures from PubChem as candidate sets.

Please mention and cite GNPS when using these data.

The implementation of the method proposed in [1] is available on:

Files description:

  • spectra.txt: informations about the MS/MS spectra (GNPS identifier, compound name and INCHI identifier)
  • data_GNPS.mat: contains the molecular fingerprints, molecular formula and InCHI corresponding to the MS/MS spectra
  • cv_ind.txt: indices of the cross-validation folds
  • ind_eval.txt: indices of the examples used for evaluation
  • candidates: fingerprints and INCHI for the different candidate sets
  • input_kernels: contains 24 input kernel matrices


[1] Brouard, C., Shen, H., Dührkop, K., d'Alché-Buc, F., Böcker, S. and Rousu, J.: Fast metabolite identification with Input Output Kernel Regression. In the proceedings of ISMB 2016, Bioinformatics 32(12): i28-i36, 2016. DOI:

[2] Dührkop, K., Shen, H., Meusel, M., Rousu, J. and Böcker, S.: Searching molecular structure databases with tandem mass spectra using CSI:FingerID. PNAS, 112(41), 12580-12585, 2015. doi:10.1073/pnas.1509788112




Files (2.9 GB)

Name Size Download all
2.9 GB Preview Download

Additional details

Related works

Is supplement to
10.1093/bioinformatics/btw246 (DOI)