Abstract: The present disclosure relates to systems, methods, and non-transitory computer-readable media that generate a modified digital image depicting a transparent object utilizing a transparency properties neural network. For example, the disclosed system accesses a trimap for a source digital image depicting a transparent object. The disclosed system utilizes the trimap of the source digital image and the source digital image to generate an alpha matte and a refractive flow. Specifically, the disclosed system generates the alpha matte and refractive flow by utilizing a transparency properties neural network. Furthermore, the disclosed system generates the modified digital image depicting the transparent object (from the source digital image) within a background of a target digital image, by modifying a portion of the background of a target digital image behind the transparent object utilizing the alpha matte and the refractive flow.

Abstract: A system and a method for converting protein data bank (PDB) files into a grayscale image array is provided. The method includes extracting PDB files from a PDB, using a data extraction module. The method also includes converting, using a file conversion module, the PDB files into a data frame based on BioPandas. The method further includes selecting, using a column selection module, one or more columns from the data frame based on a pre-determined criteria. The method furthermore includes converting, using a column conversion module, the selected one or more columns into a NumPy array for rendering the data frame to resemble an image. The method furthermore includes resizing the NumPy array using an OpenCV for making size of the NumPy array uniform and for generating a grayscale image array.

Abstract: A system and a method for converting a protein data bank file into a two-dimensional numerical matrix is provided. The method also includes extracting PDB files from a PDB, using a data extraction module. The method further includes analyzing the PDB files using an analysis module for calculating and visualizing interatomic interactions in protein structures. The method further includes generating, using a file generation module, a comma separated values (CSV) file based on the analysis. The method further includes performing a featurization of the CSV file using a featurization module. The method further includes generating, using a matrix generating module, the two-dimensional numerical matrix based on the featurization.

Abstract: A system and a method for predicting a binding affinity of protein structures based on deep learning is disclosed. The method includes capturing, using a data capture module, a multi-dimensional structure of a plurality of protein-protein complexes from a protein sequence data set. The method also includes performing featurization, using a featurization module, of the multi-dimensional structure of the plurality of protein-protein complexes by creating an adjacency matrix for successive spherical shells centered around each type of a plurality of amino acid sequences. The method further includes predicting, using a prediction module, a binding affinity from text sequence of the plurality of amino acid sequences using a pre-trained artificial intelligence model based on the adjacency matrix.

Abstract: A method and system for generating a plurality of antibody sequences of a target from one or more framework regions based on at least one model. The model is trained on a training dataset of high binding affinity to generate the complementarity determining regions (CDR) from the received one or more framework regions (FR). The generated complementarity determining regions (CDR) from the each of the one or more framework regions (FR) are combined with the associated one or more framework regions to generate one or more regions of the target. The generated one or more regions comprises each of the received one or more framework regions (FR) and corresponding each of the generated complementarity determining regions (CDR). The generated one or more regions are concatenated to generate the plurality of antibody sequences of the target. The generated plurality of antibody sequences of the target has high binding affinity.

Abstract: A system for predicting biological activities or properties of one or more chemical compounds in a new drug. The system includes a processor configured to receive a plurality of input datasets including a plurality of simplified molecular-input line-entry system (SMILES) notations and execute a pre-trained Natural Language Processing (NLP) model to transform the plurality of SMILES notations into a plurality of non-sparse matrices. The processor is configured to train a deep learning model using the plurality of non-sparse matrices to obtain a trained deep learning model, where the trained deep learning model is used to predict one or more biological activities of an untested chemical compound in the new drug when the chemical compound in the drug is subjected to at least one chemical modification to form the new drug. The system efficiently and reliably predicts the biological activities or properties of the untested chemical compound in the new drug.

Abstract: A method and system for generating a plurality of new antibody sequences corresponding to a target from a single lead antibody sequence. A model is pre-trained with a training dataset of plurality of known antibody sequences to learn a structural pattern from the plurality of known antibody sequences. The method includes receiving the single lead antibody sequence. The lead antibody sequence has one or more regions. The regions have one or more lead framework regions (FR) and one or more lead complementarity determining regions (CDR). The pre-trained model is configured to process the single lead antibody sequence to identify a relationship between the one or more lead framework regions (FR) and one or more lead complementarity determining regions (CDR). A plurality of new antibody sequences are generated from the single lead antibody amino acid of the sequence by the pre-trained model based on the identified relationship and the structural pattern.

Abstract: There is disclosed an Artificial Intelligence (AI) assisted drug discovery system and a method for generating active and inactive targets for a given drug. The generated targets comprise the complete target landscape for the given drug. The system comprises a processor which is configured to train a multi-label deep learning neural network and a predictive model to generate and validate a prediction score, and further execute the trained multi-label deep learning neural network and the trained predictive model to identify targets associated with the given drug, and generate the active and inactive targets for the given drug molecule in response to the generated prediction score for each target associated with the given drug.