Technique for extracting arrayed data

20050105787

11/018788

December 20, 2004

The current invention discloses a novel spectral transformation technique for characterizing digitized intensity output patterns from microarrays. This method yields improved sensitivity with reduced false positives and false negatives. Current microarray methods are overly sensitive to the detection of a visible distinction between pixels associated with probes and pixels associated with background. In one embodiment, a technique is disclosed that comprises the steps of: extracting pixels associated with an object of interest and transforming such pixels from an intensity representation to a spectral representation. In some embodiments, the extraction is based on a tessellated logarithmic spiral extraction that may yield a pixel core with a sampling of both foreground and background pixels. This core may then be computationally rescaled by 10×-10,000× to enhance spatial resolution. Once the extracted pixels are represented in the spectral regime, convolution with resolution-enhancement kernels may be used to accentuate morphological features capturing platform specific phenomenology.

Gulati, Sandeep (La Canada Flintridge, CA, US)

1. A method for characterizing information from a pixellated output pattern of a microarray having one or more objects of interest, comprising the steps of: extracting pixels within the output pattern representative of each object of interest using a technique chosen from the group comprising: logarithmic spiral extraction, rectilinear mask extraction, row-major extraction, and column-major extraction; and associating the extracted pixels with one or more objects of interest

2. The method of claim 1, further comprising the step of: transforming the intensity representation of the extracted pixels to a spectral representation.

3. The method of claim 2 wherein the step of transforming the intensity representations of the extracted pixels is dependent upon the expected signal level for each object of interest.

4. The method of claim 1 wherein the logarithmic spiral uses a tessellated extraction mask.

5. The method of claim 4 wherein the tessellated extraction mask is a texture mask.

6. The method of claim 5 wherein the texture mask is chosen from the group comprising: Bernoulli spiral masks, logistique masks, Hirschorn masks, Voderberg masks, bent-wedge tile masks, kinked tile masks, rhomboidal tile masks, triangular tile masks, equiangular spiral masks, symmetric tessellation masks, asymmetric tessellation masks, spiral mirabilis masks.

7. The method of claim 2 further comprising the step of computationally re-scaling the extracted pixels to increase spatial resolution of the extracted pixels prior to the step of transformation the intensity representation.

8. The method of claim 7 wherein the step of computationally re-sampling the extracted pixels comprises the step of convolving the extracted pixels with a kernel.

9. The method of claim 8 wherein the kernel is a discretized kernel.

10. The method of claim 7 wherein the step of wherein the step of computationally re-sampling the extracted pixels comprises the product of a scalar dot product with an affine transformation, or a linear function.

11. The method of claim 9 wherein the discretized coefficients of the function are chosen from the group comprising: linear functions, non-linear functions, canonical kernel functions.

12. The method of claim 7 wherein spatial resolution is increased by convolution of the extracted pixels with a canonical coefficient kernel that yields two or more modified pixels for each extracted pixel.

13. The method of claim 7 wherein the computational re-sampling of the extracted pixels utilizes a cascade of computational convolutions with discretized canonical kernel functions, with each successive convolution conducted on the results of preceding convolution.

14. The method of claim 7 wherein the convolution cascade is chosen from the group comprising: a serial cascade, a parallel cascade, or a combination of a serial cascade and a parallel cascade.

15. The method of claim 1 wherein the output pattern comprises an array of spatially contiguous pixels and the step of extracting spatial data elements further comprises the step of determining which pixels are representative of each object of interest.

16. The method of claim 1 wherein the output pattern comprises an array of spatially contiguous pixels and the step of extracting spatial data elements further comprises the step of determining which pixels associated with a morphological representation represents an object of interest.

17. The method of claim 1 wherein the output pattern comprises an array of spatially contiguous pixels and the step of extracting spatial data elements further comprises the step of determining which pixels associated with an object of interest represent a morphological invariant.

18. The method of claim 15 wherein the step of determining which pixels are representative of each object of interest is based on information-theory or an information measure.

19. The method of claim 1 further comprising transforming the domain of the extracted pixels from spatial intensity to spatial frequency.

20. The method of claim 1 wherein the step of transforming the intensity representation of the extracted pixels to a spectral representation uses a Fourier transform.

21. The method of claim 20 wherein the Fourier transform is either a Fast Fourier Transform, or a Discrete Fourier Transform.

22. The method of claim 20 further comprising the step of partitioning a spectral vector representing the transformed pixels into non-overlapping subvectors prior to convolving with discretized coefficients of a function.

23. The method of claim 20 further comprising the step of partitioning a spectral vector representing the transformed pixels into overlapping spectral vectors prior to convolving with discretized coefficients of a function.

24. The method of claim 1 further including the step of decomposing the extracted pixels into a set of discrete extracted objects, wherein each discrete extracted object is transformed into a spectral representation.

25. The method of claim 20 further comprising the step of estimating the power spectral density from the spectral representation of the extracted pixels.

26. The method of claim 20 further comprising the step of decomposing a spectrally transformed vector into subvectors associated with an object of interest.

27. The method of claim 26 wherein each subvector is convolved with a resolution enhancement kernel.

28. The method of claim 26 wherein the convolved subvectors are computationally re-sampled.

29. The method of claim 28 further comprising the step of combining the post-convolution computationally re-sampled transformed spectral subvectors to yield a single spectral vector.

30. The method of claim 1 further comprising the step of removing pixels that do not pertain to an object of interest prior to the pixel extraction step.

31. The method of claim 1 further comprising utilizing a logarithmic spiral to estimate a local background of an object of interest.

32. The method of claim 1 wherein the microarray is chosen from the group comprising: hybridized spotted cDNA microarrays, synthesized oligonucleotide arrays, spotted oligonucleotide arrays, peptide nucleotide assays, single nucleotide polymorphism (SNP) arrays, carbohydrate arrays, glycoprotein arrays, protein arrays, proteomic arrays, tissue arrays, antibody arrays, antigen arrays, bioassays, sequencing microarrays, sequencing by hybridization (SBH) microarrays, siRNA duplexes, RNAi arrays glass-based arrays, nylon membrane arrays, thin film arrays, polymer-substrate arrays, capillary electrophoresis arrays, genospectral arrays, electronic arrays, bead arrays, quantum dot arrays, and gylcan arrays.

33. The method of claim 1 wherein the output pattern is representative of an indicator chosen from the group comprising: fluorescence, chemiluminescence, bioluminescence, photoluminescence.

34. The method of claim 1 further comprising the step of registering the pixels within the output pattern to approximate the location of the objects of interest.

35. The method of claim 1 further comprising the step of adjusting the contrast of the output pattern.

36. The method of claim 35 wherein the contrast is adjusted using a filter from the group comprising: Gabor filters, low-pass band-pass filters, high-pass band-pass filters, edge detection operators, Laplacian filters, gradient-focusing filters.

37. The method of claim 1 further comprising the step of segmenting the objects of interest within the output pattern prior to the extraction step.

38. A computer code product, embodied on a computer-readable media, that characterizes information from a pixellated output pattern of microarray representing one or more objects of interest, comprising: computer code that extracts pixels within the output pattern representative of each object of interest using a technique chosen from the group comprising: logarithmic spiral extraction, rectilinear mask extraction, row-major extraction, and column-major extraction; and computer code that associates the extracted pixels with one or more objects of interest.

39. A computer system for characterizing information from a pixellated output pattern of a microarray representing one or more objects of interest, comprising: a processor; and a memory coupled to said processor, said memory encoding one or more programs, said one or more programs causing said processor to perform the following steps: extracting pixels within the output pattern representative of each object of interest using a technique chosen from the group comprising: logarithmic spiral extraction, rectilinear mask extraction, row-major extraction, and column-major extraction; and associating the extracted pixels with one or more objects of interest

40. A system for analyzing an output pattern of an array of detectors, each detector representing one or more objects of interest, comprising: means for extracting pixels within the output pattern representative of each object of interest using a technique chosen from the group comprising: logarithmic spiral extraction, rectilinear mask extraction, row-major extraction, and column-major extraction; and means associating the extracted pixels with one or more objects of interest

41. A method for characterizing information from a pixellated output pattern of a microarray having one or more objects of interest, comprising the steps of: extracting pixels within the output pattern representative of each object of interest; utilizing a logarithmic spiral to estimate a local background of an object of interest; and associating the extracted pixels with one or more objects of interest.

42. A computer code product, embodied on a computer-readable media, that characterizes information from a pixellated output pattern of microarray representing one or more objects of interest, comprising: computer code that extracts pixels within the output pattern representative of each object of interest; computer code that utilizes a logarithmic spiral to estimate a local background of an object of interest; and computer code that associates the extracted pixels with one or more objects of interest.

43. A computer system for characterizing information from a pixellated output pattern of a microarray representing one or more objects of interest, comprising: a processor; and a memory coupled to said processor, said memory encoding one or more programs, said one or more programs causing said processor to perform the following steps: extracting pixels within the output pattern representative of each object of interest; utilizing a logarithmic spiral to estimate a local background of an object of interest; and associating the extracted pixels with one or more objects of interest.

44. A system for analyzing an output pattern of an array of detectors, each detector representing one or more objects of interest, comprising: means for extracting pixels within the output pattern representative of each object of interest; means for utilizing a logarithmic spiral to estimate a local background of an object of interest; and means for associating the extracted pixels with one or more objects of interest.

382129/000

edspap.20050105787