Widespread adoption of artificial intelligence (AI) technologies is substantially affecting the human condition in ways that are not yet well understood. Negative unintended consequences abound including the perpetuation and exacerbation of societal inequalities and divisions via algorithmic decision making. We present six grand challenges for the scientific community to create AI technologies that are human-centered, that is, ethical, fair, and enhance the human condition. These grand challenges are the result of an international collaboration across academia, industry and government and represent the consensus views of a group of 26 experts in the field of human-centered artificial intelligence (HCAI). In essence, these challenges advocate for a human-centered approach to AI that (1) is centered in human well-being, (2) is designed responsibly, (3) respects privacy, (4) follows human-centered design principles, (5) is subject to appropriate governance and oversight, and (6) interacts with individuals while respecting human’s cognitive capacities. We hope that these challenges and their associated research directions serve as a call for action to conduct research and development in AI that serves as a force multiplier towards more fair, equitable and sustainable societies.
In this study, we introduce an interpretable graph-based deep learning prediction model, AttentionSiteDTI, which utilizes protein binding sites along with a self-attention mechanism to address the problem of drug–target interaction prediction. Our proposed model is inspired by sentence classification models in the field of Natural Language Processing, where the drug–target complex is treated as a sentence with relational meaning between its biochemical entities a.k.a. protein pockets and drug molecule. AttentionSiteDTI enables interpretability by identifying the protein binding sites that contribute the most toward the drug–target interaction. Results on three benchmark datasets show improved performance compared with the current state-of-the-art models. More significantly, unlike previous studies, our model shows superior performance, when tested on new proteins (i.e. high generalizability). Through multidisciplinary collaboration, we further experimentally evaluate the practical potential of our proposed approach. To achieve this, we first computationally predict the binding interactions between some candidate compounds and a target protein, then experimentally validate the binding interactions for these pairs in the laboratory. The high agreement between the computationally predicted and experimentally observed (measured) drug–target interactions illustrates the potential of our method as an effective pre-screening tool in drug repurposing applications.
With the increasing reliance on automated decision making, the issue of algorithmic fairness has gained increasing importance. In this paper, we propose a Generative Adversarial Network for tabular data generation. The model includes two phases of training. In the first phase, the model is trained to accurately generate synthetic data similar to the reference dataset. In the second phase we modify the value function to add fairness constraint, and continue training the network to generate data that is both accurate and fair. We test our results in both cases of unconstrained, and constrained fair data generation. We show that using a fairly simple architecture and applying quantile transformation of numerical attributes the model achieves promising performance. In the unconstrained case, i.e., when the model is only trained in the first phase and is only meant to generate accurate data following the same joint probability distribution of the real data, the results show that the model beats the state-of-the-art GANs proposed in the literature to produce synthetic tabular data. Furthermore, in the constrained case in which the first phase of training is followed by the second phase, we train the network and test it on four datasets studied in the fairness literature and compare our results with another state-of-the-art pre-processing method, and present the promising results that it achieves. Comparing to other studies utilizing GANs for fair data generation, our model is comparably more stable by using only one critic, and also by avoiding major problems of original GAN model, such as mode-dropping and non-convergence.
While research into drug–target interaction (DTI) prediction is fairly mature, generalizability and interpretability are not always addressed in the existing works in this field. In this paper, we propose a deep learning (DL)-based framework, called BindingSite-AugmentedDTA, which improves drug–target affinity (DTA) predictions by reducing the search space of potential-binding sites of the protein, thus making the binding affinity prediction more efficient and accurate. Our BindingSite-AugmentedDTA is highly generalizable as it can be integrated with any DL-based regression model, while it significantly improves their prediction performance. Also, unlike many existing models, our model is highly interpretable due to its architecture and self-attention mechanism, which can provide a deeper understanding of its underlying prediction mechanism by mapping attention weights back to protein-binding sites. The computational results confirm that our framework can enhance the prediction performance of seven state-of-the-art DTA prediction algorithms in terms of four widely used evaluation metrics, including concordance index, mean squared error, modified squared correlation coefficient ($r^2_m$) and the area under the precision curve. We also contribute to three benchmark drug–traget interaction datasets by including additional information on 3D structure of all proteins contained in those datasets, which include the two most commonly used datasets, namely Kiba and Davis, as well as the data from IDG-DREAM drug-kinase binding prediction challenge. Furthermore, we experimentally validate the practical potential of our proposed framework through in-lab experiments. The relatively high agreement between computationally predicted and experimentally observed binding interactions supports the potential of our framework as the next-generation pipeline for prediction models in drug repurposing.
This paper studies the interplay between social distancing and the spread of the COVID-19 disease—a global pandemic that has affected most of the world’s population. Our goals are to (1) to observe the correlation between the strictness of social distancing policies and the spread of disease and (2) to determine the optimal adoption level of social distancing policies. The earliest instances of the virus were found in China, and the virus has reached the United States with devastating consequences. Other countries severely affected by the pandemic are Brazil, Russia, the United Kingdom, Spain, India, Italy, and France. Although it is impossible to stop it, it is possible to slow down its spread to reduce its impact on the society and economy. Governments around the world have deployed various policies to reduce the virus spread in response to the pandemic. To assess the effectiveness of these policies, the system’s dynamics of the society needs to be analyzed, which is generally not possible with mathematical linear equations or Monte Carlo methods because human society is a complex adaptive system with continuous feedback loops. Because of the challenges with the other methods, we chose agent-based methods to conduct our study. Moreover, recent agent-based modeling studies for the COVID-19 pandemic show significant promise in assisting decision-makers in managing the crisis by applying policies such as social distancing, disease testing, contact tracing, home isolation, emergency hospitalization, and travel prevention to reduce infection rates. Based on modeling studies conducted in Imperial College, increasing levels of interventions could slow the spread of disease and infection. We ran the model with six different percentages of social distancing while keeping the other parameters constant. The results show that social distancing affects the spread of COVID-19 significantly, in turn decreasing the spread of disease and infection rates when implemented at higher levels. We also validated these results by using the behavior space tool with ten experiments with varying social distancing levels. We conclude that applying and increasing social distancing policy levels leads to a significant reduction in infection spread and the number of deaths. Both experiments show that infection rates are reduced drastically when social distancing intervention is implemented between 80% to 100%.
Governments, policy makers, and officials around the globe are working to mitigate the effects of the COVID-19 pandemic by making decisions that strive to save the most lives and impose the least economic costs. Making these decisions require comprehensive understanding of the dynamics by which the disease spreads. In traditional epidemiological models, individuals do not adapt their contact behavior during an epidemic, yet adaptive behavior is well documented (i.e., fear-induced social distancing). In this work we revisit Epstein’s “coupled contagion dynamics of fear and disease” model in order to extend and adapt it to explore fear-driven behavioral adaptations and their impact on efforts to combat the COVID-19 pandemic. The inclusion of contact behavior adaptation endows the resulting model with a rich dynamics that under certain conditions reproduce endogenously multiple waves of infection. We show that the model provides an appropriate test bed for different containment strategies such as: testing with contact tracing and travel restrictions. The results show that while both strategies could result in flattening the epidemic curve and a significant reduction of the maximum number of infected individuals; testing should be applied along with tracing previous contacts of the tested individuals to be effective. The results show how the curve is flattened with testing partnered with contact tracing, and the imposition of travel restrictions.