Intensity extraction for feature values in base calling
The technology disclosed extracts intensities from sequencing images for base calling target clusters and attenuates spatial crosstalk from neighboring clusters. The technology disclosed accesses a particular section from a plurality of sections of an image output by a sensor, the particular section of the image including at least one pixel depicting intensity emission values from a target cluster and neighboring clusters located across the sensor, and convolves the particular section of the image with a corresponding convolution kernel in a plurality of convolution kernels, to generate a feature map comprising a plurality of feature values. The technology disclosed further assigns a corresponding feature value to the target cluster based on feature values in the plurality of feature values adjoining a center of the target cluster, and processes the corresponding feature value assigned to the target cluster, to base call the target cluster.
1. A system comprising:
one or more processors coupled to memory; and
computer instructions that, when executed on the one or more processors, cause the system to:
access an image section of an image depicting intensity emission values from clusters of oligonucleotides;
filter, from the image section, a set of feature values for a feature map corresponding to the image section;
determine a corresponding feature value for a target cluster of oligonucleotides based on the set of feature values from the feature map; and
generate a base call for the target cluster of oligonucleotides by applying the corresponding feature value to a subset of intensity emission values for the target cluster of oligonucleotides.
2. The system of claim 1 , further comprising instructions that, when executed by the one or more processors, cause the system to determine the corresponding feature value for the target cluster of oligonucleotides by determining a weighted feature value for the target cluster of oligonucleotides based on neighboring feature values from the feature map that correspond to neighboring clusters of oligonucleotides that neighbor the target cluster of oligonucleotides.
3. The system of claim 1 , further comprising computer instructions that, when executed by the one or more processors, cause the system to:
access the image section of the image by accessing the image in an imaging channel of a sequencing instrument; and
generate the feature map for the image section in the imaging channel.
4. The system of claim 1 , further comprising computer instructions that, when executed by the one or more processors, cause the system to:
determine the corresponding feature value for the target cluster of oligonucleotides by interpolating one or more feature values from the set of feature values from the feature map for the image section of the image captured for a single sequencing cycle; and
generate the base call for the target cluster of oligonucleotides by applying the corresponding feature value to the subset of intensity emission values from the image captured for the single sequencing cycle.
5. The system of claim 1 , wherein the image section of the image comprises a sub-tile region for a sub-tile of a flow cell.
6. The system of claim 1 , further comprising instructions that, when executed by the one or more processors, cause the system to:
access an additional image section of the image depicting intensity emission values from one or more of the clusters of oligonucleotides;
filter, from the additional image section, an additional set of feature values for an additional feature map corresponding to the additional image section;
determine an additional corresponding feature value for an additional target cluster of oligonucleotides based on the additional set of feature values from the additional feature map; and
generate an additional base call for the additional target cluster of oligonucleotides by applying the additional corresponding feature value to intensity emission values for the additional target cluster of oligonucleotides.
7. The system of claim 1 , further comprising computer instructions that, when executed by the one or more processors, cause the system to filter the set of feature values for the feature map by extracting, utilizing a sharpening operation, one or more of the intensity emission values of the intensity emission values to generate one or more feature values of the set of feature values for the feature map.
8. The system of claim 1 , further comprising computer instructions that, when executed by the one or more processors, cause the system to generate the feature map by convolving the image section with a corresponding convolution kernel in a set of convolution kernels.
9. The system of claim 1 , further comprising computer instructions that, when executed by the one or more processors, cause the system to:
filter, utilizing a convolutional neural network, the set of feature values for the feature map corresponding to the image section; and
determine, utilizing the convolutional neural network, the corresponding feature value for the target cluster of oligonucleotides based on the set of feature values from the feature map.
10. A non-transitory computer readable storage medium storing computer instructions that, when executed by one or more processors, cause a system to:
access an image section of an image depicting intensity emission values from clusters of oligonucleotides;
filter, from the image section, a set of feature values for a feature map corresponding to the image section;
determine a corresponding feature value for a target cluster of oligonucleotides based on the set of feature values from the feature map; and
generate a base call for the target cluster of oligonucleotides by applying the corresponding feature value to a subset of intensity emission values for the target cluster of oligonucleotides.
11. The non-transitory computer readable storage medium of claim 10 , further storing computer instructions that, when executed by the one or more processors, cause the system to determine the corresponding feature value for the target cluster of oligonucleotides by determining a weighted feature value for the target cluster of oligonucleotides based on neighboring feature values from the feature map that correspond to neighboring clusters of oligonucleotides that neighbor the target cluster of oligonucleotides.
12. The non-transitory computer readable storage medium of claim 10 , further storing computer instructions that, when executed by the one or more processors, cause the system to:
access the image section of the image by accessing the image in an imaging channel of a sequencing instrument; and
generate the feature map for the image section in the imaging channel.
13. The non-transitory computer readable storage medium of claim 10 , further storing computer instructions that, when executed by the one or more processors, cause the system to:
determine the corresponding feature value for the target cluster of oligonucleotides by interpolating one or more feature values from the set of feature values from the feature map for the image section of the image captured for a single sequencing cycle; and
generate the base call for the target cluster of oligonucleotides by applying the corresponding feature value to the subset of intensity emission values from the image captured for the single sequencing cycle.
14. The non-transitory computer readable storage medium of claim 10 , wherein the image section of the image comprises a sub-tile region for a sub-tile of a flow cell.
15. The non-transitory computer readable storage medium of claim 10 , further storing computer instructions that, when executed by the one or more processors, cause the system to:
access an additional image section of the image depicting intensity emission values from one or more of the clusters of oligonucleotides;
filter, from the additional image section, an additional set of feature values for an additional feature map corresponding to the additional image section;
determine an additional corresponding feature value for an additional target cluster of oligonucleotides based on the additional set of feature values from the additional feature map; and
generate an additional base call for the additional target cluster of oligonucleotides by applying the additional corresponding feature value to intensity emission values for the additional target cluster of oligonucleotides.
16. A computer-implemented method comprising:
accessing an image section of an image depicting intensity emission values from clusters of oligonucleotides;
filtering, from the image section, a set of feature values for a feature map corresponding to the image section;
determining a corresponding feature value for a target cluster of oligonucleotides based on the set of feature values from the feature map; and
generating a base call for the target cluster of oligonucleotides by applying the corresponding feature value to a subset of intensity emission values for the target cluster of oligonucleotides.
17. The computer-implemented method of claim 16 , wherein filtering the set of feature values for the feature map comprises extracting, utilizing a sharpening operation, one or more of the intensity emission values of the intensity emission values to generate one or more feature values of the set of feature values for the feature map.
18. The computer-implemented method of claim 16 , further comprising generating the feature map by convolving the image section with a corresponding convolution kernel in a set of convolution kernels.
19. The computer-implemented method of claim 16 , further comprising:
filtering, utilizing a convolutional neural network, the set of feature values for the feature map corresponding to the image section; and
determining, utilizing the convolutional neural network, the corresponding feature value for the target cluster of oligonucleotides based on the set of feature values from the feature map.
20. The computer-implemented method of claim 16 , further comprising:
accessing an additional image section of the image depicting intensity emission values from one or more of the clusters of oligonucleotides;
filtering, from the additional image section, an additional set of feature values for an additional feature map corresponding to the additional image section;
determining an additional corresponding feature value for an additional target cluster of oligonucleotides based on the additional set of feature values from the additional feature map; and
generating an additional base call for the additional target cluster of oligonucleotides by applying the additional corresponding feature value to intensity emission values for the additional target cluster of oligonucleotides.