초록 ▼

Systems and methods are described that allow the high-throughput preparation, processing, and study of arrays of samples, each of which comprises at least one compound. Particular embodiments of the invention allow a large number of experiments to be performed in parallel on samples that comprised o

Systems and methods are described that allow the high-throughput preparation, processing, and study of arrays of samples, each of which comprises at least one compound. Particular embodiments of the invention allow a large number of experiments to be performed in parallel on samples that comprised of one or more compounds on the milligram or microgram quantities of compounds. Other embodiments of the invention encompass methods and devices for the rapid screening of the results of such experiments, as well as methods and devices for rapidly determining whether or not similarities exist among groups of samples in an array. Particular embodiments of the invention encompass methods and devices for the high-throughput preparation of different forms of compounds (e.g., different crystalline forms), for the discovery of new forms of old compounds, and for the discovery of new methods of producing such forms. Embodiments of the invention also allow for the high-throughput determination of how specific compounds or forms of compounds behave when exposed to other chemicals or environmental conditions.

대표청구항 ▼

1. A system for detecting similarities among a plurality of samples, which comprisesa) a device for obtaining a spectrum for each sample; and b) a computer configured to analyze each of the spectra and to generate a plurality of bins, wherein each bin corresponds to samples sharing at least one spec

1. A system for detecting similarities among a plurality of samples, which comprisesa) a device for obtaining a spectrum for each sample; and b) a computer configured to analyze each of the spectra and to generate a plurality of bins, wherein each bin corresponds to samples sharing at least one spectral feature. 2. The system of claim 1 wherein the device is an infrared spectrometer, nearinfrared spectrometer, NMR spectrometer, X-ray diffractometer, neutron diffractometer, light microscope, electron microscope, second harmonic generator, circular dichroism spectrometer, linear dichroism spectrometer, differential scanning calorimeter, thermal gravimetric analyzer, or melting point analyzer.3. The system of claim 1 wherein the device is a Raman spectrometer.4. The system of claim 1 wherein the computer is further configured to generate a binary spectral representation for a spectrum that reflects the presence or absence of a spectral feature.5. The system of claim 1 wherein the computer is configured to mutually compare a plurality of spectra and generate a hierarchical clustering dendrogram.6. The system of claim 1 wherein the computer is configured to cluster the plurality of spectra.7. The system of claim 6 wherein the computer is configured to cluster the plurality of spectra in accordance with iterative k-means clustering.8. The system of claim 1 wherein the computer is configured to cluster the plurality of spectra such that if a majority of spectra obtained from a single sample are assigned to a particular bin, then all spectra from that sample are assigned to that bin.9. The system of claim 1 wherein the computer is configured to assign newly obtained spectra to at least one of the plurality of bins.10. The system of claim 1 wherein the computer is configured to modify, in response to an analysis of newly obtained spectra, at least one of the plurality of bins.11. The system of claim 1 wherein the computer is configured to add, in response to an analysis of newly obtained spectra, at least one bin to the plurality of bins.12. The system of claim 1 wherein the computer is configured to generate a similarity matrix representing the similarity between at least two of the plurality of samples.13. The system of claim 12 wherein the computer is further configured to sort the samples such that they are arranged to reflect their similarity.14. The system of claim 12 wherein the computer is further configured to sort the similarity matrix such that a diagonal in the matrix represents samples exhibiting the greatest similarity.15. A method of detecting similarities among a plurality of samples, which comprises:a) collecting a spectrum for each of the plurality of samples; b) calculating a similarity metric between the spectrum of one sample and that of at least one other of the plurality; c) clustering, based on the similarity metric, the spectra into bins, each bin containing similar spectra; and d) presenting the clustered spectra with similar spectra located close to each other. 16. The method of claim 15 wherein the spectra are preprocessed after they are collected.17. The method of claim 16 wherein the positions of one or more spectral peaks in the preprocessed spectra are used to generate real value vectors.18. The method of claim 17 wherein binary spectra are generated from the vectors.19. The method of claim 15 wherein the spectra are Raman spectra.20. The method of claim 15, wherein said method comprises:a) providing a system for evaluating experiments comprising: i) a plurality of containers, each of which contains a compound-of-interest and optionally one or more additional compounds; ii) a block containing an array of holes for receiving the containers; and iii) an imaging device; b) positioning the block near the imaging device; c) producing images of the contents of each of the containers; d) analyzing the images for the presence of a desired experimental result; and e) identifying containers with the desired experimental result. 21. The method of claim 15, wherein said method comprises screening for solid forms of a compound-of-interest, wherein the compound-of-interest is a biologically active small organic molecule, said screening comprising: a) preparing an array of samples, each of which comprises the compound-of-interest and optionally one or more additional compounds; b) processing the array so as to generate solid forms; c) prescreening the array for solid formation using a digital imaging camera; d) identifying samples with solid formations for further analysis; e) rearranging and reprocessing samples with solids, and optionally repeating steps (b) to (d).22. The method of claim 15, wherein said method comprises screening for solid forms of a compound-of-interest, said screening comprising: a) adding a compound-of-interest and optionally one or more additional compounds into removable containers; b) placing said removable containers, before or after adding said compound-of-interest and optionally said one or more additional compounds, in a block made of a thermally conductive material having an array of holes, each hole having a top and a bottom, the top having an opening for receiving the removable container and the bottom having an access hole; c) processing said block in a thermal processing system for heating and cooling multiple blocks simultaneously, and d) and detecting said solid forms using an imaging device.23. The method of claim 15, wherein said method comprises:a) providing a system for processing a sample comprising: i) removable containers; ii) a block having an array of holes, each hole having a top and a bottom, the top having an opening for receiving a container and the bottom having an access hole; b) placing the containers in the holes; c) dispensing a controlled amount of a compound-of-interest and optionally one or more additional compounds in each container to provide an array of samples; d) processing the array; and e) screening the samples for the presence or absence of solid forms using an imaging device. 24. The method of claim 15, wherein said spectra in bins correspond to a hierarchical organization of the plurality of spectral samples based on pair-wise similarity scores calculated in accordance with said similarity metric.25. The method of claim 15, wherein said method comprises:a) providing a system which comprises: i) a block having atop surface, a bottom surface, and a plurality of holes for receiving the containers, wherein each hole has a top opening on the top surface and a bottom opening on the bottom surface, wherein the top opening is of a dimension sufficient to accommodate a container and the dimension of the bottom opening is of a dimension that will not accommodate the container; ii) a plurality of containers held in the block; and iii) a thermal processing system for heating and cooling multiple blocks simultaneously; b) dispensing a controlled amount of a compound-of-interest and optionally one or more additional compounds in each of the containers; c) optionally sealing the containers; and d) placing the block in the thermal processing system for an amount of time. 26. The method of claim 15, wherein said method comprises:a) providing a system for processing a sample comprising: i) removable containers; and ii) a block having an array of holes, each hole having a top and a bottom, the top having an opening for receiving a container and the bottom having an access hole; b) placing the containers in the holes; c) dispensing a controlled amount of a compound-of-interest and optionally one or more additional compounds in each container to provide an array of samples; d) processing the array; and e) screening the samples for the presence or absence of solid forms using an imaging device. 27. The method of claim 26, wherein said method comprises:a) imaging the contents of the containers; and b) analyzing the images for a desired experimental result. 28. The method of claim 26, wherein said method comprises: prescreening containers prior to the processing of the array wherein the prescreening further comprises:a) imaging the contents of the containers; and b) analyzing the images for a desired experimental result. 29. The method of claim 26, wherein said method comprises re-arraying the containers which contain the desired experimental results into an output block.30. The method of claim 15, using a vision station and vision algorithm to automatically detect areas of interest in a sample container comprising:a) locating or recognizing the presence or absence of a container; b) locating the meniscus, if any, of the sample in a container; and c) searching the area between the meniscus and the bottom of the container for particles, solids, solid forms, or other areas of interest. 31. The method of claim 15, comprising obtaining multiple spectra for a sample.32. The method of claim 31, comprising identifying the best spectra from a set of spectra said best spectra having the: a) highest peak signal, b) highest average signal, c) best signal to noise ratio, or d) most peaks.33. The method of claim 31, comprising constructing an average spectrum of all the spectra of a sample and using this spectrum in further processing.34. The method of claim 31, comprising constructing an agglomerated spectrum that contains the highest peak of the set for every peak window, wherein a peak window is defined as a region in which peaks are considered to be the same.35. The method of claim 31, comprising keeping all of the spectra and perform downstream analysis on all of the spectra.36. The method of claim 15, wherein hierarchical clustering is performed to represent data in the form of a similarity matrix having similar spectra listed close together.37.The method of claim 15, wherein after the spectra from all of the samples to be analyzed are collected, the spectra are processed by a series of algorithms that facilitate the binning of sample spectra according to one or more spectral features selected from: the locations of peaks, peak shoulders, peak heights, or peak areas.38. The method of claim 19, comprising eliminating artifacts of the Raman spectra that are not caused by Raman scattering.39. The method of claim 19, comprising filtering noise by a filtering technique selected from Fourier filtering, wavelet filtering, matched filtering or averaging.40. The method of claim 15, comprising finding peaks in a spectrum by finding the zero-crossings of the first derivative of a smoothed or unsmoothed spectrum, and then selecting the concave down zero-crossings that meets height and separation criteria.41. The method of claim 15, comprising creating binary spectral representations for the initial spectra.42. The method of claim 15, comprising identifying and creating cluster boundaries by a similarity measure.43. The method of claim 42, wherein the similarity measure is selected from Hamming distance, Euclidean distance or a non-similarity indices.44. The method of claim 42, wherein the similarity measure a Tversky similarity index or a derivative thereof.45. The method of claim 42, wherein the similarity measure is Tanimoto coefficient.46. The method of claim 42, wherein the similarity measure is Dice coefficient.47. A method of analyzing a plurality of samples, which comprises:a) analyzing the samples with a spectrometer to produce spectral data; b) under processor control, identifying similarities between the spectra; and c) grouping the spectra into bins of similarity. 48. The method of claim 47 wherein the spectrometer is a Raman spectrometer.49. A database containing a plurality of spectral samples organized into a plurality of bins, the bins corresponding to a hierarchical organization of the plurality of spectral samples based on pair-wise similarity scores calculated in accordance with a similarity metric.50. The database of claim 49 wherein the similarity metric is a Tanimoto coefficient, Tversky index, Euclidean distance, or Hamming distance.