This project is co-funded by the project "TEACUP: Metodi e TEcniche innovative per lo sviluppo di librerie per la modellazione, l'Analisi e il confronto CompUtazionale di Proteine", POR FSE, Programma Operativo Regione Liguria, 2014-2020, No RLOF18ASSRIC/68/1

SHREC 2021: Retrieval and classification of protein surfaces equipped with physical & chemical properties

The aim of this SHREC track is to evaluate the performance of retrieval and classification algorithms for protein surfaces characterized by physicochemical properties. Starting from a set of protein structures in different conformational states observed via NMR experiments and deposited in the PDB repository, we build their Solvent Excluded Surface (SES) by the freely available software NanoShaper. The track is jointly organized by IMATI-CNR and the CONCEPT lab at IIT. The final report of this SHREC'21 track will be submitted as a joint contribution to the international journal Computers & Graphics and will follow a two-stage review process. The paper will be authored by the track coordinators and all participants who submitted their results.

Dataset, Ground Truth and Evaluation
A dataset of approximately 5000 protein surfaces and corresponding properties will be provided. Each model will be represented by an OFF file and a TXT file. Each row of the TXT file corresponds to a vertex of the triangulation in the OFF file (in the same order); each row in the TXT file contains the physicochemical properties evaluated in the corresponding vertex in the OFF file. The dataset will be subdivided into a training and a test set (in the proportion 70%-30%). An example of the 3D models we will use in this track is shown in Figure 1. The example OFF file can be downloaded here, while the TXT file with the corresponding physicochemical properties is available here.

Each model in the test set will be used in turn as a query against the remaining part of the test set. To compare the performance of the methods equipped of physicochemical properties against the simple geometric models we will ask the participants two runs:

For a given query, the goal of the track is twofold: to retrieve the most similar objects and (optionally) to classify the query itself. The closeness of the retrieved structures with the ground truth might be evaluated a-priori on the basis of their SCOPe classification or of their sequence similarity.

Performance Evaluation: The performance of the retrieval algorithms will be evaluated using the precision-recall curves, the percentage of success for the first (PF) and the second (PS) retrieved item and the confusion matrix among the classes. For the classification task, we will adopt the classification rate and the confusion matrix.

Registration and Other Procedures
Each participant is requested to register to the track by sending an email to Andrea Raffo (email: and Ulderico Fugacci (email: with the subject "SHREC'21 track: Retrieval and classification of protein surfaces equipped with physical & chemical properties". Then, an answer will be sent to each participant with the instructions to:

We plan to submit the benchmark code and data to the replicability stamp.

Further Information

Important Dates