AlzFinder conducts virtual screening using advanced machine learning models to evaluate key targets associated with Alzheimer's disease. The platform provides the probability of a ligand's activity against the specified targets and identifies other potential targets that can be modulated. This comprehensive approach enhances the understanding of complex biological networks and aids in discovering novel therapeutic agents for Alzheimer's disease.

Users can upload files containing up to 100 compounds in various formats, including CSV, SMI, and TXT. Each line must include the SMILES code and the identification of each molecule. For example, please download the sample file available on the platform.

If you need to perform virtual screening with more than 100 compounds, please visit the AlzyFinder repository to download and install the environment and follow the provided instructions.

Users can upload a maximum of 10 molecules. Each line must contain two values separated by a space. For an example, click on the "Paste Sample" link.

Yes, the platform offers three types of results: a table with probabilities, a heat map, and a network. Each result can be downloaded in different formats: the table in CSV or PDF format, the heatmap in CSV or PNG format, and the network in PNG format.

We recommend drawing only one molecule at a time. While the molecule editor allows you to draw multiple molecules, the platform treats the entire drawing as a single molecule and calculates its probability.

If you choose not to include a name, the platform will automatically assign the label "Molecule" followed by a number based on the input order. However, we recommend naming your molecules to avoid confusion.

Yes, it is possible. To do so, you must download and install the AlzyFinder environment and follow the instructions provided in the AlzyFinder GitHub repository.

The results are going to be available on the server for 60 days.

To better understand the systemic impact that an active ligand can have on Alzheimer's disease, we took the AD proteins studied in this work and built a protein-protein interaction network (PPIn-AD) using STRING database. The PPIn-AD was expanded using NetworkX by finding the first neighbors (1st shell) in the network constructed from the initial AD proteins. With this extended PPIn-AD, users can expand the Ligand-Protein interaction network using the Expended LPIn button. As a result, proteins that interact with selected AD targets will be added to the LPIn. This extension is highly relevant, as for the first time, a web-based tool such as AlzyFinder employs machine learning to perform ligand-based virtual screening together with network pharmacology analysis to obtain information on the potential therapeutic impact of a hit molecule.

Yes, the user draws the molecules, clicks on the blue arrow, and then clicks on 'Copy as SMILES'. The molecule editor allows users to copy the structure in various formats, including mol, SMILES, InChI, InChI Key, JME, OCL, and others.

The minimum value is 0.01, and the maximum is 1 (the higher, the better). Users can select the probability according to their needs.