November 4, 2018

Theory

SoftMining offers a very wide range of products that have been developed in several European universities, and mainly at the University of Salerno.

The first thing to know is the architecture behind SoftMining. From the web interface, you can run almost all tools. The job is then sent to Grimd, a dedicated and efficient grid, that takes care of queueing and load balancing. Therefore, via a very extremely simple web interface, you can enjoy the power and the efficiency of a grid of hundreds of machines.

 

At the moment, we are porting all our tools on the new platform. This means that we can always run our calculations locally, but that if you want to enjoy directly the SoftMining capabilities you must wait a little more. We are moving tools one after the other. If you have particular needs, please contact us.

The scientific concepts of our tools:

YADA – Generally, when the binding site of a protein is known and a molecule is similar to known ligands, traditional docking methods are rather accurate in the prediction of the interaction. Unfortunately, when the binding site is unknown, the blind docking analysis requires large computational resources and the results are usually not accurate.

We have standalone Yada and a grid based Yada.

Try it here

What’s new with Yada

Yada is a new tool for molecular docking that is capable to distribute efficiently calculations onto general purposes computer grid and that combines biological and structural information of the receptor. Yada is available for Windows and Linux.
Piotto, S., Di Biasi, L., Fino, R., Parisi, R., & Sessa, L. (2016). Yada: a novel tool for molecular docking calculations. Journal of computer-aided molecular design, 30(9), 753-759.

AI_HISTO

AI_HISTO is a trained neural network for automatic categorization of histological samples of prostatic tissues. AI_HISTO must be seen as a supporting tool for prostatic cancer diagnosis. The software was developed in collaboration with PharmaBullet Srl.

Try it here

Protcomp

Protocomp is an alignment free sequence analysis tool for general purposes. Pr0tC0mp exploits compression algorithms to build up dictionaries of substrings. These dictionaries are compared by means of ad hoc metrics.

Protcomp finds numerous applications since it can be applied to any text object. We have developed Protcomp based applications for the design of novel antimicrobial peptides, for pseudo-semantic analysis of text documents, for a new semantic search engine, and even for regularity detection of electro cardiograms.