Support vector machine-based method for analysis of spectral data
View Patent ↗Support vector machines are used to classify data contained within a structured dataset such as a plurality of signals generated by a spectral analyzer. The signals are pre-processed to ensure alignment of peaks across the spectra. Similarity measures are constructed to provide a basis for comparison of pairs of samples of the signal. A support vector machine is trained to discriminate between different classes of the samples. to identify the most predictive features within the spectra. In a preferred embodiment feature selection is performed to reduce the number of features that must be considered.
1. A computer program product embodied on a non-transitory computer readable medium for predicting patterns in data contained in a plurality of spectra generated from a plurality of samples comprising a plurality of different classes, the computer program product comprising instructions for executing support vector machine classifiers and further for causing a computer processor to:
receive raw spectral data for the plurality of spectra;
align the raw spectral data for the plurality of spectra by performing the steps of:
selecting a first example spectrum as a baseline example;
sliding each spectral peak of a second example spectrum one at a time along a plurality of peaks within the baseline example;
applying a scoring function to obtain a similarity score between each spectral peak of the second example spectrum and the peaks within the baseline example; and
offsetting the second example spectrum relative to the baseline example according to the similarity score achieved for the second example spectrum;
repeating the step of aligning the spectra for at least one additional example spectrum to create a set of aligned spectra comprising aligned raw spectral data;
train and test at least one support vector machine using the set of aligned spectra to provide a trained at least one support vector machine for processing of live data;
receive raw spectral data for live spectral data;
process the live spectral data using the trained at least one support vector machine to classify the live spectral data into the plurality of different classes; and
transfer the classified live spectral data to a media.
2. The computer program product of claim 1 , wherein the baseline example is derived from the raw spectral data after applying a baseline subtraction algorithm.
3. The computer program product of claim 2 , wherein the baseline subtraction algorithm finds a smooth line that follows the bottom of a curve within the baseline spectrum without rising into the peaks.
4. The computer program product of claim 1 , wherein the similarity score is determined according to the relationship
S
(
x
i
,
x
0
)
=
∑
i
(
x
i
-
x
0
)
2
x
i
1
x
0
1
,
where x i and x 0 are feature vectors corresponding to peaks of an i th example spectrum and the baseline spectrum, respectively, and ∥x i ∥ is the l 1 norm,
x
i
1
=
∑
j
=
1
d
x
ij
.
5. The computer program product of claim 1 , wherein the similarity score is determined according to the relationship S(x i −x 0 )=∥x i, −x 0 ∥ 2 2 , where x i and x 0 are feature vectors corresponding to peaks of an i th spectrum and the baseline spectrum, respectively.
6. The computer program product of claim 1 , wherein the raw spectral data is smoothed before obtaining the similarity score.
7. The computer program product of claim 1 , wherein the raw spectral data is normalized before obtaining the similarity score.
8. The computer program product of claim 1 , wherein the plurality of spectra is mass spectrometry data.
9. The computer program product of claim 8 , wherein the plurality of spectra comprises mass spectrograms of proteins.
10. The computer program product of claim 9 , wherein the proteins are from patient samples for diagnosis of cancer.
11. The computer program product of claim 1 , wherein the data is selected from the group consisting of DNA sequences, protein sequences, documents, graphs, signals, microarray expression profiles, spectra, images, spatio-temporal data and relational data.
12. The computer program product of claim 1 , wherein the scoring of similarity is made noise invariant by discarding a lower portion of the plurality of spectra.
13. The computer program product of claim 1 , wherein the scoring of similarity is made noise invariant by using repeated measurements of the samples.
14. The computer program product of claim 1 , wherein the at least one support vector machine comprises a multi-class support vector machine.
15. The computer program product of claim 1 , further comprising instructions for causing a computer processor to, before or after aligning the raw spectral data, smooth the plurality of spectra by averaging over a window.
16. The computer program product of claim 1 , further comprising instructions for causing a computer processor to, before or after aligning the raw spectral data, extract pre-determined peaks from the plurality of spectra.
17. The computer program product of claim 1 , further comprising instructions for executing a feature selection algorithm for identifying peaks within the plurality of spectra to select a subset of spectral peaks that discriminate between the different classes, wherein the feature selection algorithm is selected from SVM-recursive feature elimination and l 0 -norm minimization.
18. The computer program product of claim 17 , further comprising instructions for generating at a printer or display device a report comprising an identification of the subset of spectral peaks.
19. The computer program product of claim 1 , further comprising instructions for generating at a printer or display device a report identifying the class corresponding to the classified live spectral data.
20. The computer program product of claim 1 , wherein the media comprises a disk drive or removable media.
21. An apparatus comprising:
a computer processor;
a memory;
a computer readable medium storing a computer program product for predicting patterns in data contained in a plurality of spectra generated from a plurality of samples comprising a plurality of different classes, the computer program product comprising instructions for executing support vector machine classifiers and further for causing a computer processor to:
receive raw spectral data for the plurality of spectra;
align the raw spectral data for the plurality of spectra by performing the steps of:
selecting a first example spectrum as a baseline example;
sliding each spectral peak of a second example spectrum one at a time along a plurality of peaks within the baseline example;
applying a scoring function to obtain a similarity score between each spectral peak of the second example spectrum and the peaks within the baseline example; and
offsetting the second example spectrum relative to the baseline example according to the similarity score achieved for the second example spectrum;
repeating the step of aligning the spectra for at least one additional example spectrum to create a set of aligned spectra comprising aligned raw spectral data;
train and test at least one support vector machine using the set of aligned spectra to provide a trained at least one support vector machine for processing of live data;
receive raw spectral data for live spectral data;
process the live spectral data using the trained at least one support vector machine to classify the live spectral data into the plurality of different classes; and
transfer the classified live spectral data to a media.