Information processing device, screening device, information processing method, screening method, and program
An information processing device includes: an acquirer configured to acquire images obtained by imaging cells differentiated from pluripotent stem cells derived from a subject; and a predictor configured to input the images acquired by the acquirer to a model trained on data in which information indicating at least a neurodegenerative disease is associated with the image obtained by imaging the cells of the neurodegenerative disease differentiated from the pluripotent stem cells, and predict an onset of the neurodegenerative disease of the subject or effects of drugs on the neurodegenerative disease, based on output results of the model to which the images were input.
1 . An information processing device comprising:
an acquirer configured to acquire images obtained by imaging cells differentiated from pluripotent stem cells derived from a subject; and
a predictor configured to:
input the images acquired by the acquirer to a neural network trained on training data in which information indicating at least a phenotype of neurodegenerative disease is associated with an image obtained by imaging cells of the neurodegenerative disease differentiated from the pluripotent stem cells, the training data used for training the neural network includes data in which at least information showing the phenotype of the neurodegenerative disease that is more prominent at a second time after a first time when the cells of the neurodegenerative disease were captured than that at the first time, is associated with the image obtained by imaging the cells of the neurodegenerative disease, and
predict an onset of the neurodegenerative disease of the subject or effects of drugs on the neurodegenerative disease based on the information indicating the phenotype output by the neural network in response to the images input.
2 . The information processing device according to claim 1 , wherein:
the neural network outputs a score indicating a likelihood that the cells showed the phenotype when the images obtained by imaging the cells were input, and
the predictor predicts the onset of the neurodegenerative disease of the subject based on the score output by the neural network.
3 . The information processing device according to claim 2 , wherein:
the training data used for training the neural network includes data in which a time period from the first time to the second time is further associated with the image obtained by imaging the cells of the neurodegenerative disease,
the neural network outputs a time period until the cells will show the phenotype from when the images obtained by imaging the cells were input, and
the predictor further predicts a time period until the onset of the neurodegenerative disease of the subject based on the time period output by the neural network.
4 . The information processing device according to claim 1 , wherein:
the training data used for training the neural network further includes data in which information that a cell does not show the phenotype of the neurodegenerative disease is associated with the image obtained by imaging the cells differentiated from the pluripotent stem cells derived from a healthy subject.
5 . The information processing device according to claim 1 , further comprising:
a learner configured to learn the neural network trained on the training data.
6 . The information processing device according to claim 5 , wherein:
the neural network includes a plurality of weak learners, and
the learner ensemble-learns the plurality of weak learners.
7 . The information processing device according to claim 6 , wherein:
the predictor inputs the images acquired by the acquirer to each of the plurality of weak learners, and predicts the onset of the neurodegenerative disease of the subject based on the information indicating the phenotype output by each of the plurality of weak learners in response to the images input.
8 . The information processing device according to claim 1 , wherein:
the cells differentiated from the pluripotent stem cells are nerve cells, glial cells, vascular endothelial cells, pericytes, choroid plexus cells, or immune system cells, and
the neurodegenerative diseases are Alzheimer's disease, Parkinson's disease, amyotrophic lateral sclerosis (ALS), spinocerebellar degeneration, frontotemporal lobar degeneration, Lewy body dementia, multiple system atrophy, Huntington's disease, progressive supranuclear palsy, or corticobasal degeneration.
9 . A screening device comprising:
an acquirer configured to acquire images obtained by imaging cells of a neurodegenerative disease, which are in contact with a test substance and differentiated from pluripotent stem cells;
a predictor configured to:
input the images acquired by the acquirer to a neural network trained on training data in which information indicating at least a phenotype of the neurodegenerative disease is associated with the image obtained by imaging the cells of the neurodegenerative disease differentiated from the pluripotent stem cells, the training data used for training the neural network includes data in which at least information showing the phenotype of the neurodegenerative disease that is more prominent at a second time after a first time when the cells of the neurodegenerative disease were captured than that at the first time, is associated with the image obtained by imaging the cells of the neurodegenerative disease, and
predict whether or not the phenotype of the neurodegenerative disease will appear in the cells of the neurodegenerative disease differentiated from the pluripotent stem cells based on the information indicating the phenotype output by the neural network in response to the images input; and
a determiner configured to determine whether the test substance is a preventive or a therapeutic agent for the neurodegenerative diseases based on prediction results of the predictor.
10 . The screening device according to claim 9 , wherein:
the determiner determines that the test substance is a preventive or a therapeutic agent for the neurodegenerative diseases in a case where the predictor predicts that the phenotype does not appear in the cells of the neurodegenerative disease differentiated from the pluripotent stem cells.
11 . An information processing method in which a computer:
acquires images obtained by imaging cells differentiated from pluripotent stem cells derived from a subject,
inputs the acquired images to a neural network trained on training data in which information indicating at least a phenotype of a neurodegenerative disease is associated with the image obtained by imaging the cells of the neurodegenerative disease differentiated from the pluripotent stem cells, the training data used for training the neural network includes data in which at least information showing the phenotype of the neurodegenerative disease that is more prominent at a second time after a first time when the cells of the neurodegenerative disease were captured than that at the first time, is associated with the image obtained by imaging the cells of the neurodegenerative disease, and
predicts an onset of the neurodegenerative disease of the subject or effects of drugs on the neurodegenerative disease based on the information indicating the phenotype output by the neural network in response to the images input.
12 . A screening method in which a computer:
acquires images obtained by imaging cells of a neurodegenerative disease, which are in contact with a test substance and differentiated from pluripotent stem cells,
inputs the acquired images to a neural network trained on training data in which information indicating at least a phenotype of the neurodegenerative disease is associated with the image obtained by imaging the cells of the neurodegenerative disease differentiated from the pluripotent stem cells, the training data used for training the neural network includes data in which at least information showing the phenotype of the neurodegenerative disease that is more prominent at a second time after a first time when the cells of the neurodegenerative disease were captured than that at the first time, is associated with the image obtained by imaging the cells of the neurodegenerative disease,
predicts whether or not the phenotype of the neurodegenerative disease will appear in the cells of the neurodegenerative disease differentiated from the pluripotent stem cells based on the information indicating the phenotype output by the neural network in response to the images input, and
determines whether the test substance is a preventive or a therapeutic agent for the neurodegenerative diseases based on prediction results.
13 . A computer-readable non-transitory storage medium storing a program causing a computer to execute:
acquires images obtained by imaging cells differentiated from pluripotent stem cells derived from a subject,
inputs the acquired images to a neural network trained on training data in which information indicating at least a phenotype of a neurodegenerative disease is associated with the image obtained by imaging the cells of the neurodegenerative disease differentiated from the pluripotent stem cells, the training data used for training the neural network includes data in which at least information showing the phenotype of the neurodegenerative disease that is more prominent at a second time after a first time when the cells of the neurodegenerative disease were captured than that at the first time, is associated with the image obtained by imaging the cells of the neurodegenerative disease, and
predicts an onset of the neurodegenerative disease of the subject or effects of drugs on the neurodegenerative disease based on the information indicating the phenotype output by the neural network in response to the images input.
14 . A computer-readable non-transitory storage medium storing a program causing a computer to execute:
acquires images obtained by imaging cells of a neurodegenerative disease, which are in contact with a test substance and differentiated from pluripotent stem cells,
inputs the acquired images to a neural network trained on training data in which information indicating at least a phenotype of the neurodegenerative disease is associated with the image obtained by imaging the cells of the neurodegenerative disease differentiated from the pluripotent stem cells, the training data used for training the neural network includes data in which at least information showing the phenotype of the neurodegenerative disease that is more prominent at a second time after a first time when the cells of the neurodegenerative disease were captured than that at the first time, is associated with the image obtained by imaging the cells of the neurodegenerative disease,
predicts whether or not the phenotype of the neurodegenerative disease will appear in the cells of the neurodegenerative disease differentiated from the pluripotent stem cells based on the information indicating the phenotype output by the neural network in response to the images input, and
determines whether the test substance is a preventive or a therapeutic agent for the neurodegenerative diseases based on prediction results.