DTNPD: A comprehensive database of drugs and targets for neurological and psychiatric disorders

In response to the shortcomings in data quality and coverage for neurological and psychiatric disorders (NPDs) in existing comprehensive databases, this paper introduces the DTNPD database, specifically designed for NPDs. DTNPD contains detailed information on 30 NPDs types, 1847 drugs, 514 drug targets, 64 drug combinations, and 61 potential target combinations, forming a network with 2389 drug-target associations. The database is user-friendly, offering open access and downloadable data, which is crucial for network pharmacology studies. The key strength of DTNPD lies in its robust networks of drug and target combinations, as well as drug-target networks, facilitating research and development in the field of NPDs. The development of the DTNPD database marks a significant milestone in understanding and treating NPDs. For accessing the DTNPD database, the primary URL is http://dtnpd.cnsdrug.com, complemented by a mirror site available at http://dtnpd.lyhbio.com.

Drug-target interaction prediction with tree-ensemble learning and output space reconstruction

Background

Computational prediction of drug-target interactions (DTI) is vital for drug discovery. The experimental identification of interactions between drugs and target proteins is very onerous. Modern technologies have mitigated the problem, leveraging the development of new drugs. However, drug development remains extremely expensive and time consuming. Therefore, in silico DTI predictions based on machine learning can alleviate the burdensome task of drug development. Many machine learning approaches have been proposed over the years for DTI prediction. Nevertheless, prediction accuracy and efficiency are persisting problems that still need to be tackled. Here, we propose a new learning method which addresses DTI prediction as a multi-output prediction task by learning ensembles of multi-output bi-clustering trees (eBICT) on reconstructed networks. In our setting, the nodes of a DTI network (drugs and proteins) are represented by features (background information). The interactions between the nodes of a DTI network are modeled as an interaction matrix and compose the output space in our problem. The proposed approach integrates background information from both drug and target protein spaces into the same global network framework.

Results

We performed an empirical evaluation, comparing the proposed approach to state of the art DTI prediction methods and demonstrated the effectiveness of the proposed approach in different prediction settings. For evaluation purposes, we used several benchmark datasets that represent drug-protein networks. We show that output space reconstruction can boost the predictive performance of tree-ensemble learning methods, yielding more accurate DTI predictions.

Conclusions

We proposed a new DTI prediction method where bi-clustering trees are built on reconstructed networks. Building tree-ensemble learning models with output space reconstruction leads to superior prediction results, while preserving the advantages of tree-ensembles, such as scalability, interpretability and inductive setting.

Systems psychopharmacology: A network approach to developing novel therapies

The multifactorial origin of most chronic disorders of the brain, including schizophrenia, has been well accepted. Consequently, pharmacotherapy would require multi-targeted strategies. This contrasts to the majority of drug therapies used until now, addressing more or less specifically only one target molecule. Nevertheless, quite some searches for multiple molecular targets specific for mental disorders have been undertaken. For example, genome-wide association studies have been conducted to discover new target genes of disease. Unfortunately, these attempts have not fulfilled the great hopes they have started with. Polypharmacology and network pharmacology approaches of drug treatment endeavor to abandon the one-drug one-target thinking. To this end, most approaches set out to investigate network topologies searching for modules, endowed with "important" nodes, such as "hubs" or "bottlenecks", encompassing features of disease networks, and being useful as tentative targets of drug therapies. This kind of research appears to be very promising. However, blocking or inhibiting "important" targets may easily result in destruction of network integrity. Therefore, it is suggested here to study functions of nodes with lower centrality for more subtle impact on network behavior. Targeting multiple nodes with low impact on network integrity by drugs with multiple activities ("dirty drugs") or by several drugs, simultaneously, avoids to disrupt network integrity and may reset deviant dynamics of disease. Natural products typically display multi target functions and therefore could help to identify useful biological targets. Hence, future efforts should consider to combine drug-target networks with target-disease networks using mathematical (graph theoretical) tools, which could help to develop new therapeutic strategies in long-term psychiatric disorders.