Abstract: Amino acid sequences of proteins can be produced using an autoencoder. For example, amino acid sequences of variant proteins can be produced by an autoencoder that is fed an amino acid sequence of a base protein as input. A decoding component of the autoencoder can include at least one or more components of a generative adversarial network.

Abstract: A system for generating a model for predicting a molecular property of a variant of a molecule is provided. For each of a plurality of variants of the molecule, the system for each structural feature, aggregates the values for the structural features of the residues of the molecule that were modified to form the variant to form a feature vector for the variant. The system assigns the value for the molecular property of the variant to the feature vector wherein the feature vector and the assigned value form training data. The system then generates the model for predicting a value for the molecular property using the training data for the plurality of variants.

Abstract: Amino acid sequences of proteins can be produced using an autoencoder. For example, amino acid sequences of variant proteins can be produced by an autoencoder that is fed an amino acid sequence of a base protein as input. A decoding component of the autoencoder can include at least one or more components of a generative adversarial network.

Abstract: Amino acid sequences of proteins can be produced using one or more generative machine learning architectures. The amino acid sequences produced by the one or more generative machine learning architectures can be used to train a classification model architecture. The classification model architecture can classify amino acid sequences according to a number of classifications. Individual classifications of the number of classifications can correspond to at least one of a structural feature of proteins, a range of values of a structural feature of proteins, a biophysical property of proteins, or a range of values of a biophysical property of proteins.

Abstract: Amino acid sequences of antibodies can be generated using a generative adversarial network that includes a first generating component that generates amino acid sequences of antibody light chains and a second generating component that generates amino acid sequences of antibody heavy chains. Amino acid sequences of antibodies call be produced by combining the respective amino acid sequences produced by the first generating component and the second generating component. The training of the first generating component and the second generating component can proceed at different rates. Additionally, the antibody amino acids produced by combining amino acid sequences front the first generating component and the second generating component may be evaluated according to complentarity-determining regions of the antibody amino acid sequences.

Abstract: Amino acid sequences of antibodies can be generated using a generative adversarial network that includes a first generating component that generates amino acid sequences of antibody light chains and a second generating component generates amino acid sequences of antibody heavy chains. Amino acid sequences of antibodies can be produced by combining the respective amino acid sequences produced by the first generating component and the second generating component. The training of the first generating component and the second generating component can proceed at different rates. Additionally, the antibody amino acids produced by combining amino acid sequences from the first generating component and the second generating component may be evaluated according to complentarity-determining regions of the antibody amino acid sequences.

Abstract: Systems and techniques are described to generate amino acid sequences of target proteins based on amino acid sequences of template proteins using machine learning techniques. The amino acid sequences of the target proteins can be generated based on data that constrains the modifications that can be made to the amino acid sequences of the template proteins. In illustrative examples, the template proteins can include antibodies produced by a non-human mammal that bind to an antigen and the target proteins can correspond to human antibodies with a region having at least a threshold amount of identity with the binding region of the template antibody. Generative adversarial networks can be used to produce the amino acid sequences of the target proteins.

Abstract: The concepts described herein are directed to implementations of production facilities that can produce molecules used to treat biological conditions, such as biotherapeutics. The biotherapeutics can include various molecules, such as proteins, enzymes, and antibodies. The production facilities can include a number of separate modular cleanrooms that comprise particular pieces of equipment to perform one or more aspects of the processes used to manufacture biotherapeutics. The modular cleanrooms are arranged such that material that is produced by the equipment of one modular cleanroom can be transferred to another modular cleanroom for additional processing. Additionally, systems and processes are described to generate models using machine learning techniques, where the models can be used to predict productivity and/or efficiency metrics for production lines of biotherapeutics. Further, models can be generated to control the operation of pieces of equipment included in the production lines.

Abstract: Amino acid sequences of antibodies can be generated using a generative adversarial network that includes a first generating component that generates amino acid sequences of antibody light chains and a second generating component generates amino acid sequences of antibody heavy chains. Amino acid sequences of antibodies can be produced by combining the respective amino acid sequences produced by the first generating component and the second generating component. The training of the first generating component and the second generating component can proceed at different rates. Additionally, the antibody amino acids produced by combining amino acid sequences from the first generating component and the second generating component may be evaluated according to complentarity-determining regions of the antibody amino acid sequences.

Abstract: Technologies are described related to determining protein structure and properties based on sequences of proteins. In various implementations, a first model can be generated to determine structural features of proteins based on amino acid sequences of the proteins. Additionally, a second model can be generated to determine biophysical properties of proteins based on structural features of the proteins. In particular implementations, an amino acid sequence of a particular protein can be utilized by the first model to determine one or more structural features of the protein. The one or more structural features of the protein generated by the first model can be utilized by the second model to determine at least one biophysical property of the protein.

Abstract: Technologies are described related to determining conditions for the purification of proteins. In some implementations, models can be generated that predict yield and purity for various chromatographic techniques at a number of pH levels and salt concentrations. Training data used to produce the models can be obtained by running chromatography columns using stationary phase materials of different chromatographic techniques at various combinations of pH levels and salt concentrations. In additional implementations, the training data can be obtained by analyzing solutions included in a subset of wells of a multi-well plate, where the subset of wells are associated with particular salt concentrations and pH values for a particular stationary phase material. Further, the models can be used to determine optimized conditions for the purification of various proteins based on maximizing yield and purity, while minimizing cost.

Abstract: Technologies are described related to determining the impact of substitutions of amino acid sequences of proteins on properties of the base protein. Values of properties for proteins that include a particular substitution are analyzed with respect to values of properties for proteins that do not include the particular substitution. The analysis can be utilized to determine the impact of the particular substitution on the properties of the proteins while minimizing the number of proteins that need to be expressed. The impact of the particular substitution on the proteins can indicate changes to the stability and/or yield of the proteins.

Abstract: A system for generating a model for predicting a molecular property of a variant of a molecule is provided. For each of a plurality of variants of the molecule, the system for each structural feature, aggregates the values for the structural features of the residues of the molecule that were modified to form the variant to form a feature vector for the variant. The system assigns the value for the molecular property of the variant to the feature vector wherein the feature vector and the assigned value form training data. The system then generates the model for predicting a value for the molecular property using the training data for the plurality of variants.