Welcome to project Pre-alignment
Biomarker discovery based on profiling of complex body fluids requires analysis of large number of samples and demands LC-MS(/MS) platforms with high sample throughput. The most challenging step in processing such type of data is accurate correction of non-linear retention time shifts between chromatograms. There are two conditions to align successfully a pair of LC-MS (/MS) chromatograms:
1) the samples should contain minimal number of common compounds
2) peak elution order of common compounds should be the same.
Current time alignment methods do not test if these conditions are met, and cannot report to users if it is possible to align two chromatograms. We have developed a novel quality control method to assess if two LC-MS(/MS) chromatograms can be aligned by testing these two conditions. Figure on the right contains the most important step of this quality control method. The first step consists of matching peaks in two chromatograms with suitable accuracy. Finding common peaks in two chromatograms is performed by identifying peak pairs either by agreement of sequence of identified peptides or matching peaks in two chromatograms that are close in mass and in intensity rank order of de-isotoped and charge de-convoluted peak lists. This is followed by the separation of the accurately matched peak pairs from inaccurately matched ones by using 2-dimensional kernel density estimation (2DKS). When more than 100 accurately matched peak pairs are found, LOWESS regression is used to calculate a monotonic retention time correspondence function. Finally the probability of peak elution order similarity is determined by comparing the orthogonal residual variances of the accurately matched peak pairs along the retention time trend obtained between the two pairs of chromatograms: one pair being the two chromatograms of interest and the second two chromatograms that have the same elution order of common peaks. If no significant peak order inversion is detected, the retention time alignment correspondence function is used to align the retention time of peaks in the sample peak lists (peak list in red on the right figure).
The source code is available for SVN checkout, anonymous readonly subversion access works as follows:
svn co https://trac.nbic.nl/svn/pre-alignment pre-alignment
You can download the zipped code file from - https://trac.nbic.nl/pre-alignment/downloads section.
Please check the information about how to use the tool and for detailed explanation of the input / output arguments in Usage page
Other NBIC software projects
All active NBIC software projects can be accessed from the project index.
For comments or queries please go to - https://trac.nbic.nl/pre-alignment/newticket