Hexabind

Hexabind is an AI-based, high-speed, and highly accurate tool for predicting binding affinity between proteins and small molecules.

📄️ User Inputs

The user input area is shown in Figure 1. The user inputs the protein-ligand pairs for which binding affinities are to be predicted in one of three different ways (Figure 1, “1”). If selecting the “Fasta” or “PDB ID” radio button options, a single .csv file is to be uploaded via the “Browse” button (Figure 1, “2”). This file should have a column for the ligand SMILES (Figure 1, “3”), and a column for the FASTA sequence or the PDB ID of the protein (Figure 1, “4”), depending on the radio button option chosen. The “Download Sample” option (above and to the right of the “Browse” button) provides an example of the input file. Upon specifying the SMILES and FASTA/PDB ID columns, a preview of the first 5 rows of the columns is provided (Figure 1, “5”) as a quick check for whether the correct columns have been selected. Descriptive tags can be entered for the job (Figure 1, “6”) for easy identification of the job in the jobs list (see below). Note: The user needs to hit “enter” for the job tags to be applied. The job is submitted via the “submit” button (Figure 1, “7”).

📄️ Job Submission

Upon submission, the job is added to the submitted jobs list with a message acknowledging the successful submission (green ribbon in Figure 2). Completed jobs have a status of either “SUCCESS” or “FAILURE”, while jobs in progress have a “RECEIVED” status. The refresh icon at the top right of the jobs table can be used to check for completion of a “RECEIVED” job.

📄️ Results

Upon successful completion of a job, the “Visualize Results” and “Download Results” options become available in the jobs table (Figure 2). The results table displayed upon clicking the “Visualize Results” option is shown in Figure 3. Besides the table of calculated binding affinities for each protein-ligand pair, the screen also provides filtering and display options. A drop-down menu is available upon clicking the “Columns” button, where the user can specify the columns to be displayed in the table. Each of these properties can also be used as sorting and filtering criteria. Sorting can be performed by clicking the arrow button that appears upon hovering over any column header, and the filtering function can be accessed via the drop-down menu in the “Filters” button. The height of the rows can be adjusted via the “Density” button, and the “Export” button provides the option to download the properties table as a “.csv” file.