TranDTA: Prediction Of Drug Target Binding Affinity Using Transformer Representations

Drug discovery is generally difficult, expensive and the success rate is low. One of the essential steps in the early stages of drug discovery and drug repurposing is identifying drug target interactions. Although several methods developed use binary classification to predict if the interaction between a drug and its target exists or not, it is more informative and challenging to predict the strength of the binding between a drug and its target. Binding affinity indicates the strength of drug-target pair interactions. In this regard, several computational methods have been developed to predict the drug-target binding affinity. With the advent of deep learning methods, the accuracy of binding affinity prediction is improving. However, the input representation of these models is very effective in the result. The early models only use the sequence of molecules and the latter models focus on the structure of them. Although the recent models predict binding affinity more accurate than the first ones, they need more data and resources for training. In this study, we present a method that uses a pre-trained transformer to represent the protein as model input. Although pretrained transformer extracts a feature vector of the protein sequence, they can learn structural information in layers and heads. So, the extracted feature vector by transformer includes the sequence and structural properties of protein. Therefore, our method can also be run without limitations on resources (memory, CPU and GPU). The results show that our model achieves a competitive performance with the state-of-art models. Data and trained model is available at http://bioinformatics.aut.ac.ir/TranDTA/ .

MACSima Imaging Cyclic Staining (MICS) Technology Reveals Combinatorial Target Pairs for CAR T Treatment of Solid Tumors

Many critical advances in research utilize techniques that combine high-resolution with high-content characterization at the single cell level. We introduce the MICS (MACSimaTM Imaging Cyclic Staining) technology, which enables the immunofluorescent imaging of hundreds of protein targets across a single specimen at sub-cellular resolution. MICS is based on cycles of staining, imaging, and erasure, using photobleaching of fluorescent labels of recombinant antibodies (REAfinityTM), release of antibodies (REAleaseTM) or their labels (REAdyeleaseTM). Multimarker analysis can identify potential targets for immune therapy against solid tumors. With MICS we analysed human glioblastoma, ovarian and pancreatic carcinoma, and 16 normal tissues. One potential target pair for chimeric antigen receptor (CAR) T-cell therapies identified for ovarian carcinoma is EPCAM/THY1. Using an adapter CAR T cell approach, we show selective killing of cells only in presence of both markers. MICS represents a new high content microscopy methodology to be widely used for personalized medicine.

Efficient representations of tumor diversity with paired DNA-RNA aberrations

Cancer cells display massive dysregulation of key regulatory pathways due to now well-catalogued mutations and other DNA-related aberrations. Moreover, enormous heterogeneity has been commonly observed in the identity, frequency and location of these aberrations across individuals with the same cancer type or subtype, and this variation naturally propagates to the transcriptome, resulting in myriad types of dysregulated gene expression programs. Many have argued that a more integrative and quantitative analysis of heterogeneity of DNA and RNA molecular profiles may be necessary for designing more systematic explorations of alternative therapies and improving predictive accuracy. We introduce a representation of multi-omics profiles which is sufficiently rich to account for observed heterogeneity and support the construction of quantitative, integrated, metrics of variation. Starting from the network of interactions existing in Reactome, we build a library of “paired DNA-RNA aberrations” that represent prototypical and recurrent patterns of dysregulation in cancer; each two-gene “Source-Target Pair” (STP) consists of a “source” regulatory gene and a “target” gene whose expression is plausibly “controlled” by the source gene. The STP is then “aberrant” in a joint DNA-RNA profile if the source gene is DNA-aberrant (e.g., mutated, deleted, or duplicated), and the downstream target gene is “RNA-aberrant”, meaning its expression level is outside the normal, baseline range. With M STPs, each sample profile has exactly one of the 2M possible configurations. We concentrate on subsets of STPs, and the corresponding reduced configurations, by selecting tissue-dependent minimal coverings, defined as the smallest family of STPs with the property that every sample in the considered population displays at least one aberrant STP within that family. These minimal coverings can be computed with integer programming. Given such a covering, a natural measure of cross-sample diversity is the extent to which the particular aberrant STPs composing a covering vary from sample to sample; this variability is captured by the entropy of the distribution over configurations. We apply this program to data from TCGA for six distinct tumor types (breast, prostate, lung, colon, liver, and kidney cancer). This enables an efficient simplification of the complex landscape observed in cancer populations, resulting in the identification of novel signatures of molecular alterations which are not detected with frequency-based criteria. Estimates of cancer heterogeneity across tumor phenotypes reveals a stable pattern: entropy increases with disease severity. This framework is then well-suited to accommodate the expanding complexity of cancer genomes and epigenomes emerging from large consortia projects.

108 Attentional Control Deficits during Total Sleep Deprivation: Independence from Reduced Vigilant Attention

Introduction Total sleep deprivation (TSD) has been shown to impair performance on a two-phase attentional control task, the AX-type continuous performance task with switch (AX-CPTs). Here we investigate whether the observed AX-CPTs impairments are a downstream consequence of TSD-induced non-specific effects (e.g., reduced vigilant attention) or reflect a distinct impact on attentional control. Methods N=55 healthy adults (aged 26.0±0.7y; 32 women) participated in a 4-day laboratory study with 10h baseline sleep (22:00-08:00) followed by 38h TSD and then 10h recovery sleep. At baseline (09:00 day 2) and after 25h and 30h TSD (09:00 and 14:00 day 3), subjects were tested on a 10min psychomotor vigilance test (PVT), an assay of vigilant attention, and on the AX-CPTs. The AX-CPTs required subjects to differentiate designated target from non-target cue-probe pairs. In phase 1, target trials occurred frequently, which promoted prepotent anticipatory responses; in phase 2, the target pair was switched. Accuracy of responses to various different AX-CPTs trial types was expressed relative to accuracy on phase 1 neutral (non-target cue and probe) trials, which should capture non-specific impairments on the task. For all three test sessions, these relative accuracy measures, along with accuracy on phase 1 neutral trials and lapses (RT>500ms) on the PVT, were subjected to principal component analysis (PCA). Results The PCA revealed three statistically independent factors. Following varimax rotation, factor 1 (36.3% variance explained) and factor 3 (14.8% variance explained) each had high loadings for relative accuracy on multiple AX-CPTs trial types from phases 1 and 2; whereas factor 2 (17.9% variance explained) had high loadings for accuracy on phase 1 neutral trials, relative accuracy on phase 1 target trials, and PVT lapses. Conclusion These results indicate a statistical separation between AX-CPTs phase 1 neutral trials and phase 1 target trials, in conjunction with PVT lapses, versus the various other AX-CPTs trial types. This suggests a dissociation between TSD-induced, non-specific impairments on the task—potentially related to reduced vigilant attention—and TSD-induced specific impairments related to attentional control. Thus, TSD-induced deficits in attentional control are unlikely to be a downstream consequence of non-specific impairments. Support (if any) CDMRP grant W81XWH-16-1-0319

A Geometric Deep Learning Approach to Predict Binding Conformations of Bioactive Molecules

Understanding the interactions formed between a ligand and its molecular target is key to guide the optimization of molecules. Different experimental and computational methods have been key to understand better these intermolecular interactions. Herein, we report a method based on geometric deep learning that is capable of predicting the binding conformations of ligands to protein targets. Concretely, the model learns a statistical potential based on distance likelihood which is tailor-made for each ligand-target pair. This potential can be coupled with global optimization algorithms to reproduce experimental binding conformations of ligands. We show that the potential based on distance likelihood described in this paper performs similar or better than well-established scoring functions for docking and screening tasks. Overall, this method represents an example of how artificial intelligence can be used to improve structure-based drug design.

A Geometric Deep Learning Approach to Predict Binding Conformations of Bioactive Molecules

Understanding the interactions formed between a ligand and its molecular target is key to guide the optimization of molecules. Different experimental and computational methods have been key to understand better these intermolecular interactions. Herein, we report a method based on geometric deep learning that is capable of predicting the binding conformations of ligands to protein targets. Concretely, the model learns a statistical potential based on distance likelihood which is tailor-made for each ligand-target pair. This potential can be coupled with global optimization algorithms to reproduce experimental binding conformations of ligands. We show that the potential based on distance likelihood described in this paper performs similar or better than well-established scoring functions for docking and screening tasks. Overall, this method represents an example of how artificial intelligence can be used to improve structure-based drug design.

Feature Based Method for Predicting Pharmacological Interaction

Prediction of drug target interaction is an extrusive domain of drug discovery and repositioning of drugs. Most conventional studies are carried out in early years in the wet laboratory, but it is very expensive and time consuming. So nowadays, the use of machine learning techniques to predict drug target pairs. A new method of interaction targeting drugs is introduced in this paper. Use the Pseudo Position Specific Scoring Matrix (PsePSSM) is used to represent the target, which generate features that describe the original information of protein. The drug chemical structure information can be extracted through FP2 molecular fingerprint which describe the molecular structure information. Then a drug target interaction network is constructed using bipartite graph where in which each node represents a target or drug and each link indicates a drug target interaction. From the above stages, the data contains some noise and redundant data which have a negative impact on the prediction output. So, LASSO (Least Absolute Shrinkage and Selection Operator) method is handle it and reduce the dimension of the extracted feature information of original data. But drug target pair samples have some imbalanced, then cost-sensitive ensemble method is used to address the imbalanced problem between positive and negative samples, and learns about the minority class by assigning higher costs and optimizing their cost error. Finally, the processed data is given as input to the extreme gradient boosting classifier algorithm for predicting new drug target interaction pairs. This method can significantly improve the prediction accuracy of drug target interaction.

DTI-MLCD: predicting drug-target interactions using multi-label learning with community detection method

Identifying drug-target interactions (DTIs) is an important step for drug discovery and drug repositioning. To reduce the experimental cost, a large number of computational approaches have been proposed for this task. The machine learning-based models, especially binary classification models, have been developed to predict whether a drug-target pair interacts or not. However, there is still much room for improvement in the performance of current methods. Multi-label learning can overcome some difficulties caused by single-label learning in order to improve the predictive performance. The key challenge faced by multi-label learning is the exponential-sized output space, and considering label correlations can help to overcome this challenge. In this paper, we facilitate multi-label classification by introducing community detection methods for DTI prediction, named DTI-MLCD. Moreover, we updated the gold standard data set by adding 15,000 more positive DTI samples in comparison to the data set, which has widely been used by most of previously published DTI prediction methods since 2008. The proposed DTI-MLCD is applied to both data sets, demonstrating its superiority over other machine learning methods and several existing methods. The data sets and source code of this study are freely available at https://github.com/a96123155/DTI-MLCD.