We have updated the site to 1.0.1. CSI:FingerID is now trained on GNPS 1.2.8 as well as on Orbitrap and Q-TOF spectra from Massbank. The candidate list now contains a simple filter function to only display compound candidates that are contained in some bio databases. There will be further updates within the next days and weeks.
February 2nd, 2016

Our paper for CSI:FingerID has appeared! See Proceedings of the National Academy of Sciences USA, doi:10.1073/pnas.1509788112, 2015.
September 22nd, 2015

• Kai Dührkop, Huibin Shen, Marvin Meusel, Juho Rousu, and Sebastian Böcker Searching molecular structure databases with tandem mass spectra using CSI:FingerID PNAS 2015 ; published ahead of print September 21, 2015, doi:10.1073/pnas.1509788112
• Huibin Shen, Kai Dührkop, Sebastian Böcker and Juho Rousu. Metabolite Identification through Multiple Kernel Learning on Fragmentation Trees. Bioinformatics, 30(12):i157-i164, 2014. Proc. of Intelligent Systems for Molecular Biology (ISMB 2014).

• H. Horai, M. Arita, S. Kanaya, Y. Nihei, T. Ikeda, K. Suwa. Y. Ojima, K. Tanaka, S. Tanaka, K. Aoshima, Y. Oda, Y. Kakazu, M. Kusano, T. Tohge, F. Matsuda, Y. Sawada, M. Yokota Hirai, H. Nakanishi, K. Ikeda, N. Akimoto, T. Maoka, H. Takahashi, T. Ara, N. Sakurai, H. Suzuki, D. Shibata, S. Neumann, T. Iida, K. Tanaka, K. Funatsu, F. Matsuura, T. Soga, R. Taguchi, K. Saito and T. Nishioka MassBank: A public repository for sharing mass spectral data for life sciences J. Mass Spectrom., 45, 703-714 (2010).

• GNPS (4879 spectra)
• Massbank (2755 spectra)

How is CSI:FingerID working?

Given tandem mass spectra of a compound, CSI:FingerID first predicts its molecular fingerprint using Machine Learning. Then, it uses the predicted fingerprint to search in a molecular structure database, to find a best matching structure.

What is a molecular fingerprint?

Molecular fingerprints encode the structure of a molecule in a binary vector (string), where each position (digit) represents the presence or absence of a particular substructure in the molecule.

How is CSI:FingerID different from FingerID?

CSI:FingerID integrates fragmentation trees into the analysis pipeline and, doing so, reaches a strongly improved search performance. See the PNAS paper for an evaluation.

People have been using molecular fingerprints for decades! What is the fuss?

Molecular fingerprints were previously used to screen molecular structures. Predicting molecular fingerprints from tandem mass spectrometry data was pioneered by Heinonen et al. in 2012.
Heinonen M, Shen H, Zamboni N, Rousu J.Metabolite identification and molecular fingerprint prediction through machine learning.Bioinformatics 28(18):2333-41, 2012.)

What is the correct fragmentation energy?

The best thing probably is to mix fragmentation spectra at different energies, such as 10, 20, and 40 eV. Simply paste the spectra one after another.

What mass accuracy is required?

As a rule-of-thumb, 10 ppm or better should be sufficient. Be reminded of the difference between "mass accuracy" and "anecdotal mass accuracy" (Zubarev R, Mann M.On the proper use of mass accuracy in proteomics.Mol Cell Proteomics 6(3):377-81, 2007). In detail, we expect that 95% to 99% of all measured peaks have mass accuracy 10 ppm.

What instrument type is required?

CSI:FingerID has been trained on QTOF data only. It should work for any type of CID fragmentation data, though.

How do I upload RAW files?

Sorry, you cannot. The software would not know which feature to process, anyways.

How can I process compounds in batch mode?

Currently, there is no batch mode.

Can I get a copy of your MS data?

GNPS and MassBank are open repositories, please download from there.

What other possibilities do I have to analyze my data?

You should probably first search openly available spectral libraries. For searching in momolecular structure databases, there is CFM-ID, MAGMa, and MetFrag.

CSI:FingerID (Compound Structure Identification: FingerID) is a method for searching a tandem mass spectrum of a small molecule (metabolite) in a database of molecular structures. Details on the method can be found in the publications:

Novel versions of CSI:FingerID will become available in the near future; in particular, we will retrain the method when more training is available.