Predictor#
The Predictor API provided by OpenProtein.AI allows you to train property predictors from your assays, building on top of embeddings and reductions (e.g. SVD) provided by our Embeddings API.
You can list the available embeddings models that can be used with protein language model (plm) feature types using the /embeddings/models endpoint. Do note that, currently, using PLMs will require a reduction type to be specified (e.g. MEAN or SUM).
Similarly, SVD embeddings can also be specified using svd feature type with the appropriate fitted svd, which can be listed using the /svd endpoint.
Currently, we support the following predictor types:
Gaussian Process