Methods and systems for determining the biological age of samples
The present disclosure provides systems, media, and methods for applying machine learning to determine the biological age of a sample. A method of using a biological age predictor may comprise: (a) providing data of a sample to the biological age predictor; (b) treating the sample with a substance, thereby generating a treated sample; and (c) providing data of the treated sample to the biological age predictor. The biological age predictor may generate the first biological age and/or the second biological age.
1. A method comprising:
(a) obtaining a biological sample;
(b) obtaining first computer-readable image assay data of said biological sample, said first computer-readable image assay data generated by imaging said biological sample, wherein said first computer-readable image assay data identifies cell morphology or cellular components of said biological sample;
(c) providing a computer processor programmed with a biological age predictor, wherein said biological age predictor comprises a trained machine learning detection structure configured to receive as inputs computer-readable image assay data generated by imaging test biological samples and to determine as outputs biological ages or ranges of biological ages of said test biological samples, wherein said trained machine learning detection structure is trained using (i) computer-readable image assay data comprising images of training biological samples generated by imaging said training biological samples, wherein said computer-readable image assay data identifies cell morphology or cellular components of said training biological samples, and (ii) biological ages or ranges of biological ages associated with said plurality of training biological samples;
(d) using said computer processor programmed with said biological age predictor, determining (i) a biological age of said biological sample or (ii) a range of said biological age of said biological sample based on said first computer-readable image assay data;
(e) contacting said biological sample with a substance, thereby generating a treated biological sample;
(f) obtaining second computer-readable image assay data of said treated biological sample, said second computer-readable image assay data generated by imaging said treated biological sample, wherein said second computer-readable image assay data identifies cell morphology or cellular components of said treated biological sample; and
(g) using said computer processor programmed with said biological age predictor, determining (i) a biological age of said treated biological sample or (ii) a range of said biological age of said treated biological sample based on said second computer-readable image assay data.
2. The method of claim 1 , wherein said biological sample is a cell sample.
3. The method of claim 2 , wherein said cell sample is a skin, muscle, blood, liver, heart, spleen, thymus, brain, myoblast, fibroblast, or keratinocyte cell sample.
4. The method of claim 1 , wherein said biological sample is a tissue sample.
5. The method of claim 4 , wherein said tissue sample is a skin, fibroblast, muscle, liver, heart, spleen, thymus, or brain tissue sample.
6. The method of claim 1 , wherein said trained machine learning detection structure comprises a neural network, a convolutional neural network (CNN), a deep convolutional neural network (DCNN), a cascaded deep convolutional neural network, a simplified CNN, a shallow CNN, or a combination thereof.
7. The method of claim 1 , wherein said first computer-readable image assay data or said second computer-readable image assay data identifies biomarkers of said biological sample or said treated biological sample.
8. The method of claim 1 , wherein said substance is an anti-aging agent, a drug entity, or an FDA-approved drug entity.
9. The method of claim 1 , further comprising determining said biological age of said biological sample, or said biological age of said treated biological sample.
10. The method of claim 1 , further comprising determining said range of said biological age of said biological sample, or said range of said biological age of said treated biological sample.
11. A method comprising:
(a) obtaining a biological sample;
(b) performing an image assay on said biological sample to generate computer-readable image assay data of said biological sample, wherein said computer-readable image assay data identifies cell morphology or cellular components of said biological sample;
(c) providing a computer processor programmed with a biological age predictor, wherein said biological age predictor comprises a trained machine learning detection structure configured to receive as inputs computer-readable image assay data generated by imaging test biological samples and to determine as outputs biological ages or ranges of biological ages of said test biological samples, wherein said trained machine learning detection structure is trained using (i) computer-readable image assay data comprising images of training biological samples generated by imaging said training biological samples, wherein said computer-readable image assay data identifies cell morphology or cellular components of said training biological samples, and (ii) biological ages or ranges of biological ages associated with said plurality of training biological samples;
(d) using said computer processor programmed with said biological age predictor, analyzing a set of one or more locations in said computer-readable image assay data to:
(i) identify a feature of said set of one or more locations and classify said feature, and
(ii) determine (i) a biological age of said biological sample or (ii) a range of said biological age of said biological sample based on said identified and classified feature.
12. The method of claim 11 , wherein said biological sample is a cell sample.
13. The method of claim 12 , wherein said cell sample is a skin, muscle, blood, liver, heart, spleen, thymus, brain, myoblast, fibroblast, or keratinocyte cell sample.
14. The method of claim 11 , wherein said biological sample is a tissue sample.
15. The method of claim 14 , wherein said tissue sample is a skin, fibroblast, muscle, blood, liver, heart, spleen, thymus, or brain tissue sample.
16. The method of claim 11 , wherein said computer-readable image assay data identifies biomarkers of said biological sample.
17. The method of claim 11 , further comprising (i) determining said biological age of said biological sample.
18. The method of claim 11 , further comprising (ii) determining said range of said biological age of said biological sample.
19. A system comprising: a digital processing device comprising:
(a) a computer processor programmed with a biological age predictor, wherein said biological age predictor comprises a trained machine learning detection structure configured to receive as inputs computer-readable image assay data generated by imaging test biological samples and to determine as outputs biological ages or ranges of biological ages of said test biological samples, wherein said trained machine learning detection structure is trained using (i) computer-readable image assay data comprising images of training biological samples generated by imaging said training biological samples, wherein said computer-readable image assay data identifies cell morphology or cellular components of said training biological samples, and (ii) biological ages or ranges of biological ages associated with said plurality of training biological samples; and
(c) a computer program including instructions executable by said computer processor to create an application comprising:
a biological age assessment module configured to use said biological age predictor to analyze computer-readable image assay data of a biological sample, said computer-readable image assay data generated by imaging said biological sample, wherein said first computer-readable image assay data identifies cell morphology or cellular components of said biological sample, wherein said trained machine learning detection structure of said biological age predictor analyzes a set of one or more locations in said computer-readable image assay data to (1) identify a feature of said set of one or more locations and classify said feature, and (2) determine (i) a biological age of said biological sample or (ii) a range of said biological age of said biological sample based on said identified and classified feature.
20. The system of claim 19 , wherein said trained machine learning detection structure comprises a neural network, a convolutional neural network (CNN), a deep convolutional neural network (DCNN), a cascaded deep convolutional neural network, a simplified CNN, a shallow CNN, or a combination thereof.
21. The system of claim 19 , wherein said computer-readable image assay data identifies biomarkers of said biological sample.
22. The system of claim 19 , wherein said biological age assessment module determines said biological age of said biological sample.
23. The system of claim 19 , wherein said biological age assessment module determines said range of said biological age of said biological sample.
24. A method for screening a test compound to identify an anti-aging agent, comprising:
(a) contacting a biological sample with said test compound, thereby generating a treated biological sample;
(b) performing an image assay on said treated biological sample to generate computer-readable image assay data of said treated biological sample, wherein said computer-readable image assay data identifies cell morphology or cellular components of said treated biological sample;
(c) providing a computer processor programmed with a biological age predictor, wherein said biological age predictor comprises a trained machine learning detection structure configured to receive as inputs computer-readable image assay data generated by imaging test biological samples and to determine as outputs biological ages or ranges of biological ages of said test biological samples, wherein said trained machine learning detection structure is trained using (i) computer-readable image assay data comprising images of training biological samples generated by imaging said training biological samples, wherein said computer-readable image assay data identifies cell morphology or cellular components of said training biological samples, and (ii) biological ages or ranges of biological ages associated with said plurality of training biological samples;
(d) using said computer processor programmed with said biological age predictor, determining (i) a biological age of said treated biological sample or (ii) a range of said biological age of said treated biological sample; and
(e) identifying said test compound as an anti-aging agent based on (i) said biological age of said treated biological sample or (ii) said range of said biological age of said treated biological sample.
25. The method of claim 1 , wherein contacting said biological sample with said substance is performed in vitro.
26. The method of claim 11 , wherein said trained machine learning detection structure comprises a neural network, a convolutional neural network (CNN), a deep convolutional neural network (DCNN), a cascaded deep convolutional neural network, a simplified CNN, a shallow CNN, or a combination thereof.
27. The method of claim 26 , wherein said trained machine learning detection structure comprises a CNN.
28. The method of claim 24 , wherein said trained machine learning detection structure comprises a neural network, a convolutional neural network (CNN), a deep convolutional neural network (DCNN), a cascaded deep convolutional neural network, a simplified CNN, a shallow CNN, or a combination thereof.
29. The method of claim 24 , wherein said computer-readable image assay data identifies biomarkers of said treated biological sample.
30. The method of claim 24 , wherein contacting said biological sample with said test compound is performed in vitro.