We develop a pipeline to mine complex drug interactions by combining different similarities and interaction types (molecular, structural, phenotypic, genomic etc). Our goal is to learn an optimal kernel from these heterogeneous similarities in a supervised manner. We formulate an extensible framework that can easily integrate new interaction types into a rich model. The core of our pipeline features a novel kernel-learning approach that tunes the weights of the heterogeneous similarities, and fuses them into a Similarity-based Kernel for Identifying Drug-Drug interactions and Discovery, or SKID3. Experimental evaluation on the DrugBank database shows that SKID3 effectively combines similarities generated from chemical reaction pathways (which generally improve precision) and molecular and structural fingerprints (which generally improve recall) into a single kernel that gets the best of both worlds, and consequently demonstrates the best performance.

No abstract

Counting the number of true instances of a clause is arguably a major bottleneck in relational probabilistic inference and learning. We approximate counts in two steps: (1) transform the fully grounded relational model to a large hypergraph, and partially-instantiated clauses to hypergraph motifs; (2) since the expected counts of the motifs are provably the clause counts, approximate them using summary statistics (in/outdegrees, edge counts, etc). Our experimental results demonstrate the efficiency of these approximations, which can be applied to many complex statistical relational models, and can be significantly faster than state-of-the-art, both for inference and learning, without sacrificing effectiveness.

No abstract

The essential oil (EO) composition of the aerial parts of Erigeron multiradiatus (Lindl.ex DC.) Benth growing wild in the central Himalayan region of Uttarakhand, India, was analyzed by capillary gas chromatography with a flame ionization detector and gas chromatography–mass spectrometry. A sum of 12 constituents was identified, representing 97.81% of the oil composition. The oil was composed mainly of oxygenated monoterpenes (88.95%), sesquiterpene hydrocarbons (5.61%), oxygenated sesquiterpenes (3.05%), and monoterpene hydrocarbons (0.20%). Major constituents identified were trans-2-cis-8-matricaria-ester (77.79%), cis-lachnophyllum ester (11.04%), zingiberene (4.43%), and spathulenol (1.59%). Further, the leishmanicidal effect of EO and the purified compound trans-2-cis-8-matricaria-ester has been investigated against Leishmania donovani promastigotes and intracellular amastigotes. EO and trans-2-cis-8-matricaria-ester were safer for the hamster peritoneal macrophage and lethal to promastigotes and intracellular amastigotes at different concentrations. Further, using an in silico approach, these four compounds were tested against 10 major proteins of L. donovani associated with its virulence. Out of them, only trans-2-cis-8-matricaria-ester was found to be effective against the four target proteins, namely, l-asparaginase-1-like protein, metacaspase 2, metacaspase 1, and DNA topoisomerase II of L. donovani. The results indicate that EO contains trans-2-cis-8-matricaria-ester as a major component and showed antileishmanial activity which may facilitate discovery of new lead molecules for developing herbal medicines against visceral leishmaniasis.

No abstract

The goal of combining the robustness of neural networks and the expressivity of symbolic methods has rekindled the interest in Neuro-Symbolic AI. One specifically interesting branch of research is deep probabilistic programming languages (DPPLs) which carry out probabilistic logical programming via the probability estimations of deep neural networks. However, recent SOTA DPPL approaches allow only for limited conditional probabilistic queries and do not offer the power of true joint probability estimation. In our work, we propose an easy integration of tractable probabilistic inference within a DPPL. To this end we introduce SLASH, a novel DPPL that consists of Neural-Probabilistic Predicates (NPPs) and a logical program, united via answer set programming. NPPs are a novel design principle allowing for the unification of all deep model types and combinations thereof to be represented as a single probabilistic predicate. In this context, we introduce a novel +/- notation for answering various types of probabilistic queries by adjusting the atom notations of a predicate. We evaluate SLASH on the benchmark task of MNIST addition as well as novel tasks for DPPLs such as missing data prediction, generative learning and set prediction with state-of-the-art performance, thereby showing the effectiveness and generality of our method.

No abstract