Method of extracting locations of nucleic acid array features
AssigneeAgilent Technologies, Inc.
InventorsHerbert F. Cattell, Andreas N. Dorsel, John W. Sadler, Nicholas M. Sampas
Life of this patent
Estimating the term…
Abstract
Methods for correcting systematic errors in the measured position of deposited features of a nucleic acid array on a substrate. Systematic errors are modeled by an algorithmic model based on measuring the positions (and possibly other properties) of a subset of the features, and a model is constructed for predicting deviations in feature position from an ideal grid. Deviations arising in the deposition process, the scanning process, or both may be corrected.
Claims5
A method of measuring feature intensities of features of a nucleic acid array on a substrate comprising:
- determining size of features and uncertainty in feature placement from ideal feature locations, based on known properties of at least one of a deposition system used to deposit said features and a scanning system used to scan said features;
- using the determined size and uncertainty to calculate the size of the smallest area which is known to contain an entire feature;
- and
- measuring intensity of an observable property for an area having the calculated size and centered on the ideal location of a feature to determine an intensity for that feature.
A method of selecting a subset of a group of physical features of a nucleic acid array on a substrate for construction of an algorithmic model of feature location, comprising:
- determining a set of ideal feature locations where the physical features would be deposited on the substrate if no measurable errors existed in a deposition system used to deposit the physical features on the substrate;
- scanning the substrate to generate a set of pixel data corresponding to the intensity of an observable property in the vicinity of each ideal feature location;
- determining whether the pixel data for each ideal feature location define a strong feature by applying a criterion determined by one or more properties of the pixel data selected from the group consisting of pixel magnitude, number of pixels having magnitudes above a threshold value, locations of pixels having magnitudes above a threshold value, and distribution of pixel magnitude values;
- and
- selecting the set of ideal feature locations determined by the criterion to define strong features.
Dependencies are inferred from the claim text.