NIST Peptide Mass Spectral Libraries

:: Biochemical Reference Data for Protein and Proteome Analysis

peptide MS/MS spectrum

Goal of This Project

To build comprehensive, 'high-quality' libraries of peptide ion fragmentation mass spectra (MS/MS).

Purpose of This Site

Provide download access to the libraries.
Provide download access to the latest NIST software for use with the libraries.
Provide links to other related projects and collaborators.

Project Description

Peptide fragmentation mass spectra are used to identify peptides, and by inference proteins in biological samples. These spectra are commonly produced in large numbers by researchers in the field of proteomics using mass spectrometers to catalog (or sequence) proteins present in a sample. Equally as common are experiments designed to compare two or more samples for differentially expressed proteins. An example of the latter would be a so-called "biomarker discovery" experiment, in which the differentially expressed proteins (e.g. expressed in a cancer sample but not a healthy control) might be pursued as candidates with diagnostic or therapeutic potential.

Currently, interpretation of peptide fragmentation mass spectra is accomplished using numerous computer algorithms which rely on protein sequence to generate a theoretical spectrum for comparison. While this approach has worked well in the absence of libraries of empirical spectra, the goal of this project is to develop such libraries for all biologically relevant peptides, thereby providing a common basis for interpretation of unknown peptide mass spectra and improving the reproducibility of those interpretations.

Background Summary

This project was initiated in 2004 as an offshoot of the NIST/EPA/NIH Mass Spectral Library, a widely used product for the identification of small molecules by GC-MS analysis. The underlying parallels between an EI library and a peptide MS/MS library make them amenable to similar data analysis methods and workflows. The contents of these libraries have been derived from the interpretation of millions of peptide tandem mass spectra from many sources.

The primary motivation for this project was to enable the advantages of spectral library searching for peptide identification (see advantages below). It is hypothesized that improving the sensitivity and robustness of peptide spectrum interpretation by computer algorithms will reduce variability, leading to a reduction in both false positive and false negative identifications.

The rapid development of the peptide libraries is enabled by:

The predictable fragmentation of peptides in a mass spectrometer.
The availability of peptide tandem mass from community and internal sources.
Computer algorithms that automate the interpretation of the mass spectra.

The use of a spectral library for identification of peptides has advantages over current sequence-based search methods.

Speed because the reference spectra are known a priori thus removing the need to calculate them from protein sequence before matching. The search space is also limited to previously identified peptides, reducing the number of candidate spectra per query.
'Spectrum-spectrum' comparisons can be more discriminating, leading to improved separation of true and false matches. This is due in part to the use of intensity values (i.e., peak heights), the occurence of non-cannonical peaks, and the use of spectrum annotations during scoring.

Intended Use

This collection is intended primarily to demonstrate the utility of peptide ion fragmentation libraries by enabling the development of software applications that use them. However, in some cases (E. coli, human and yeast) the libraries are extensive enough to be employed for practical use. Emphasis has been placed on quality control, since an error in a library can be widely propagated. For example, library spectra containing significant impurity peaks can falsely identify the impurity in an analysis as the library peptide. On the other hand, to maximize completeness and minimize false negative results (peptide ions not in library), extensive extraction methods were employed. These methods combined results of several sequence search engines for initial peptide identification. Spectra were annotated in detail to aid spectrum library scoring as well as to document the origin of the spectrum. It is important to note that none of these collections are complete. Additional measurements are needed to increase both the coverage and the quality of the library.

Get involved!

This is a community-based program, and all of the peptide libraries are distributed free of charge. If you wish to contribute spectra to an existing library or suggest we build a new library, drop us a line. At the moment, we are seeking donations of raw data from all human sample types, as well as data from model organisms.

Peptide Mass Spectral Libraries (Rel. 3.0, Latest)

Be sure to download the library corresponding to the instrument type you are using.

Copyright information: "These peptide mass spectral libraries are protected by copyright law and may not be re-distributed without a valid Distribution Agreement. To receive such an agreement, contact the Standard Reference Data Program at the National Institute of Standards and Technology by email data at nist.gov or call 301-975-2008."

How to reference these libraries.

Build Date	Download*			Seq. File†	Library Browser^NEW	Library	Organism	Instrument	Exps.‡	Spectra	Peptides	Coverage
By species
Feb. 04, 2009	ASCII	NIST	SpectraST	FASTA	Browse	human	H. sapiens	it	348	261,778	158,522	16%
Jul. 22, 2008	ASCII	NIST	SpectraST	FASTA	Browse	human	H. sapiens	qtof	61	12,473	10,139	1%
Apr. 29, 2009	ASCII	NIST	SpectraST	FASTA	Browse	mouse	M. musculus	it	117	131,628	78,613	8%
Jul. 14, 2008	ASCII	NIST	SpectraST	FASTA	Browse	drosophila	D. melanogaster	it	97	96,542	62,162	13%
Jun. 26, 2009	ASCII	NIST	SpectraST	FASTA	Browse	C.elegans	C.elegans	it	2	81,177	49,754	12%
May 04, 2009	ASCII	NIST	SpectraST	FASTA	Browse	yeast	S. cerevisiae	it	63	87,676	52,076	23%
May 06, 2009	ASCII	NIST	SpectraST	FASTA	Browse	yeast	S. cerevisiae	qtof	5	3,176	2,960	2%
May 21, 2009	ASCII	NIST	SpectraST	FASTA	Browse	E. coli	E. coli	it	42	54,479	32,480	23%
May 21, 2009	ASCII	NIST	SpectraST	FASTA	Browse	rat	R. norvegicus	it	25	20,992	15,206	2%
Individual Protein Standards and Mixtures
Sept. 26, 2008	ASCII	NIST	SpectraST	FASTA	Browse	NCI20	Std. Proteins	it	301	3,903	1,892	35%
Apr. 15, 2009	ASCII	NIST	SpectraST	FASTA	Browse	chicken (egg)	G. gallus	it	42	4,472	2,387	0%
Jul. 09, 2008	ASCII	NIST	SpectraST	FASTA	Browse	Sigma UPS1	Std. Proteins	it	20	3,542	1,838	89%
May 06, 2009	ASCII	NIST	SpectraST	FASTA	Browse	serum albumin	H. sapiens	it	26	2,487	1,099	100%
May 06, 2009	ASCII	NIST	SpectraST	FASTA	Browse	BSA	B. tarus	it	8	969	458	99%
May 06, 2009	ASCII	NIST	SpectraST	FASTA	Browse	beta-2-micoglobulin	H. sapiens	it	2	179	122	100%
Jul. 15, 2008	ASCII	NIST	SpectraST	FASTA	Browse	aurum	Std. Proteins	tof-tof	1	596	545	9%
May 06, 2009	ASCII	NIST	SpectraST	FASTA	Browse	c-reactive protein	H. sapiens	it	1	94	57	77%

*See the 'Search Engines' section below for selecting the appropriate format for your software.
^†FASTA (protein sequence) files used to construct the spectral libraries. These are for REFERENCE ONLY.
^‡An experiment is often represented by many LC-MS/MS data files.

A note on modifications: Post-translational modifications (or biological PTMs) and other chemical derivatives of peptides have been limited to the following (PSI-MS names) in the above mass spectral libraries:

Carbamidomethylation of C
Oxidation of M
Acetylation of the N-terminus
Pyro-carbamidomethyl
Glu->pyro-Glu
Gln->pyro-Glu
ICAT (several forms, if applicable)
Carbamylation of the N-terminus (if applicable)

We are currently evaluating candidate peptide spectra containing additional mods for the libraries.

Documentation and Technical Information

Information on library construction and annotation is available by request.
Growth of Selected Libraries

Search Engines and Software

NIST

NIST MSPepSearch^NEW - Developed by the Chemical Reference Data Group at NIST. This program is a console implementation of Nistdl32a.dll (available by request) with numerous improvements for searching large peptide libraries. Scoring is similar to that described here and here. The program handles input spectra in MGF format, and results are written to a tab-delimited output file. Use with NIST-formatted libraries. Currently available only for Windows^TM. **Download**
NIST MS Search 2.0 (Demo) - A demo version of the graphical user interface program MS Search 2.0 for searching and viewing spectrum matches. Also allows browsing/viewing of library spectra. Input spectra in MGF or MSP format are read by the program. Use with NIST-formatted libraries. **Download**
ReAdW4Mascot2.exe - A version of ReAdW.exe (pronounced "read raw") developed at ISB that has been modified at NIST to extract and calculate additional values from Thermo^TM .RAW files. This file converts a RAW file to MGF and/or mzXML and requires an installation of Thermo's XCalibur^©. This program and its output fields will be described in an upcoming publication. **Download**
NISTMSQC^NEW - A new software pipeline for monitoring LC-MS/MS performance. This software can be used to quantitate run-to-run and/or lab-to-lab variability in one or more "shotgun" proteomics data sets. This work was first developed with funding from the NCI's CPTAC initative. **Link**

The Institute for Systems Biology

SpectraST - [PMID: 17295354 and 18806791] Developed by Henry Lam while in the Aebersold Group at ISB. Henry is now at Hong Kong University of Science and Technology. SpectraST can be download as part of the Trans Proteomic Pipeline (TPP). A user's Wiki is also available. Use with SpectraST-formatted libraries. **Download**

MacCoss Lab at the University of Washington

BiblioSpec - [PMID: 16906711]. A NIST to BiblioSpec library conversion tool is under development. In the meantime, NIST libraries are not useable with BiblioSpec. **Link**.
Skyline^NEW - An application for choosing SRM transitions for targeted proteomics applications. Spectral libraries are used as reference for this application. Use with ASCII-formatted (MSP) libraries. **Download**

The Global Proteome Machine

X!Hunter - [PMID: 16889405] developed by Craig and Beavis and part of the The GPM. A NIST to X!Hunter library conversion tool is also under development. **Link**

About Us

The NIST Mass Spectrometry Data Center develops software and databases for the identification of chemical compounds by mass spectrometry.
The center is located on the NIST main campus in Gaithersburg, MD and is a part of the Biochemical and Chemical Reference Data Division.
We collaborate with many other government, academic and commercial entities in an effort to better provide 'high-quality' reference data and software products.

NIST statement on privacy, security, and accessibility.

Inquires concerning this site and/or its contents should be addressed to paul.rudnick@nist.gov.