NMR spectroscopy can provide a wide range of information for metabolic characterization of biological samples. However, the quantity and complexity of spectroscopic data obtained from metabolomic studies has made data interpretation very difficult. The challenges stem not only from the need to apply standard processing procedures to large sets of spectra and connect them with auxiliary information about the samples, but also the frequent necessity to invoke pattern recognition (PR) methods to ensure optimal information retrieval. Most existing software deals only with a few of the aspects of this process and in general acts on a single spectrum.

Here we present a software tool, High Resolution Spectroscopy (HiRes), to address all these challenges and provide comprehensive analysis of metabolic datasets. It combines standard spectral processing routines, data correction functions, techniques for reducing information complexity of multi-spectral dataset such as Principal Component Analysis (PCA), and tools for PR, such as Non-Negative Matrix Factorization (NMF), which extracts a set of spectral features with direct physical interpretation that can be used for identifying meaningful biochemical effects. These spectral features are also correlated with auxiliary information from the samples, such as biological end-points, drug toxicity or disease diagnosis.

HiRes is provided as a free software program for research and clinical purposes. It is written in C++ and runs on Windows platform. It has been funded by Metabolic roadmap initiative of National Institutes of Health DK070301.

• Data Importation and Organization:
  • Automatic importation of NMR spectra acquired from different instruments such as Bruker and Varian spectrometers.
  • Other data type can be imported by manually entering parameters.
  • Series of spectra are imported in a single step, together with any additional information associated with the data, such as drug dose, measurement time, from a separate tab-delimited text file.
  • All the spectra and associated information are organized in a project and listed in a table that can be easily sorted by each column.
• Spectral Data Processing:
  • Data processing procedures can be applied either to all the spectra or selected individual spectra.
  • Constant phase and linear phase correction.
  • Automatic phasing.
  • Temporal filtering ( Lorentzian or Gaussian).
  • Baseline correction.
  • Water suppression.
  • PPM origin adjustment.
  • Zero fill or truncate FID to user specified length.
  • Normalization methods using acquisition number, single reference peak, or entire spectrum.
  • Peak alignment using either highest point or Principal Component Analysis (PCA).
  • Spectral binning.
• Spectral Pattern Recognition:
  • Principal Component Analysis (PCA) to identify variation sources in the data set.
  • Non-negative Matrix Factorization (NMF) to retrieve spectral patterns with meaningful biological effects.
  • Spectral integration with user-controlled baseline correction.
• Data Output and Program Extensibility:
  • Various result output options, such as spectral export in text or binary format, analysis result export to text file, printing, etc.
  • Interface for users to extend the functionality on the basis of HiRes.

You can also download a demo project here. The demo project comprises a series of NMR spectra from rat urine samples after administration of a toxin at various doses, contributed by Imperial College London. The spectra has been processed using HiRes, such as phasing, baseline correction, peak align, etc. First uncompress it, and then load hires_demo.tpj to HiRes program using "open project" menu item. Then you can try out all the functionalities provided by HiRes.

The program will expire six months after the first run on your computer. Please feel free to contact Qi Zhao to get a newer version. If you process your data using HiRes, we would appreciate your acknowledgement of the software and sending us a copy of your publication.

HiRes is an ongoing project, with new features constantly added. Your feedback would be highly appreciated and beneficial to the software's continuous improvement. Please send your questions, comments or bug reports to Qi Zhao at qz2106@columbia.edu. We look forward to hearing from you!

• New update v1.5 ---- 11/27/2007
  Some of the changes are (for more details, refer to the help document):
  1. Added spectral binning.
  2. Improved peak alignment.
  3. Bug fixes.
• Update v1.4 ---- 05/16/2007
  Some of the new additions are (for more details, refer to the help document):
  1. Added 3-D display of the results from PCA and cNMF.
  2. Improved local peak alignment by the edge smoothing.
  3. Added capability to zero fill or truncate FID to user specified length.
  4. Added support for more data types.
  5. Bug fixes.
• Update v1.2 ---- 09/29/2006
  Some of the new additions are (for more details, refer to the help document):
  1. Added spectral integration with user-controlled baseline correction.
  2. Added selected spectral list, so that procedures can applied to selected spectra only.
  3. Added initial peak alignment by singlet or doublet.
  4. Added support for multiple peak region selection for PCA and NMF.
  5. Added support for loading spectral properties after project is created.
  6. Added support for more data types, such as processed Bruker spectra, multi-fid Varian,etc.
• First release v1.0 ---- 03/05/2006