Qatar Cardiovascular Research Center

Computational techniques for predicting molecular properties are emerging as key components for streamlining drug development, optimizing time and financial investments. Here, we introduce ChemLM, a transformer language model for this task. ChemLM leverages self-supervised domain adaptation on chemical molecules to enhance its predictive performance. Within the framework of ChemLM, chemical compounds are conceptualized as sentences composed of distinct chemical ‘words’, which are employed for training a specialized chemical language model. On the standard benchmark datasets, ChemLM either matched or surpassed the performance of current state-of-the-art methods. Furthermore, we evaluated the effectiveness of ChemLM in identifying highly potent pathoblockers targeting Pseudomonas aeruginosa (PA), a pathogen that has shown an increased prevalence of multidrug-resistant strains and has been identified as a critical priority for the development of new medications. ChemLM demonstrated substantially higher accuracy in identifying highly potent pathoblockers against PA when compared to state-of-the-art approaches. An intrinsic evaluation demonstrated the consistency of the chemical language model’s representation concerning chemical properties. The results from benchmarking, experimental data and intrinsic analysis of the ChemLM space confirm the wide applicability of ChemLM for enhancing molecular property prediction within the chemical domain.

Retrieval-Augmented Generation (RAG) has been shown to enhance the factual accuracy of Large Language Models (LLMs), but existing methods often suffer from limited reasoning capabilities in effectively using the retrieved evidence, particularly when using open-source LLMs. To mitigate this gap, we introduce a novel framework, OPEN-RAG, designed to enhance reasoning capabilities in RAG with opensource LLMs. Our framework transforms an arbitrary dense LLM into a parameter-efficient sparse mixture of experts (MoE) model capable of handling complex reasoning tasks, including both single-and multi-hop queries. OPEN-RAG uniquely trains the model to navigate challenging distractors that appear relevant but are misleading. As a result, OPEN-RAG leverages latent learning, dynamically selecting relevant experts and integrating external knowledge effectively for more accurate and contextually relevant responses. In addition, we propose a hybrid adaptive retrieval method to determine retrieval necessity and balance the trade-off between performance gain and inference speed. Experimental results show that the Llama2-7Bbased OPEN-RAG outperforms state-of-the-art LLMs and RAG models such as ChatGPT, Self-RAG, and Command R+ in various knowledgeintensive tasks. We open-source our code and models at https://openragmoe.github.io/

Data visualization is a powerful tool for understating information through visual cues. However, allowing novices to create visualization artifacts for what they want to see is not easy, just as not everyone can write SQL queries. Arguably, the most natural way to specify what to visualize is through natural language or speech, similar to our daily search on Google or Apple Siri, leaving to the system the task of reasoning about what to visualize and how.

We model the behavioral biases of human decision-making in securing interdependent systems and show that such behavioral decision-making leads to a suboptimal pattern of resource allocation compared to non-behavioral (rational) decision-making. We provide empirical evidence for the existence of such behavioral bias model through a controlled subject study with 145 participants. We then propose three learning techniques for enhancing decision-making in multi-round setups. We illustrate the benefits of our decision-making model through multiple interdependent real-world systems and quantify the level of gain compared to the case in which the defenders are behavioral. We also show the benefit of our learning techniques against different attack models. We identify the effects of different system parameters on the degree of suboptimality of security outcomes due to behavioral decision-making.

With the outbreak of the COVID-19 pandemic, people turned to social media to read and to share timely information including statistics, warnings, advice, and inspirational stories. Unfortunately, alongside all this useful information, there was also a new blending of medical and political misinformation and disinformation, which gave rise to the first global infodemic. While fighting this infodemic is typically thought of in terms of factuality, the problem is much broader as malicious content includes not only fake news, rumors, and conspiracy theories, but also promotion of fake cures, panic, racism, xenophobia, and mistrust in the authorities, among others. This is a complex problem that needs a holistic approach combining the perspectives of journalists, fact-checkers, policymakers, government entities, social media platforms, and society as a whole. With this in mind, we define an annotation schema and detailed annotation instructions that reflect these perspectives. We further deploy a multilingual annotation platform, and we issue a call to arms to the research community and beyond to join the fight by supporting our crowdsourcing annotation efforts. We perform initial annotations using the annotation schema, and our initial experiments demonstrated sizable improvements over the baselines.

The time-critical analysis of social media streams is important for humanitarian organizations to plan rapid response during disasters. The crisis informatics research community has developed several techniques and systems to process and classify big crisis-related data posted on social media. However, due to the dispersed nature of the datasets used in the literature, it is not possible to compare the results and measure the progress made towards better models for crisis informatics. In this work, we attempt to bridge this gap by combining various existing crisis-related datasets. We consolidate eight annotated data sources and provide 166.1k and 141.5k tweets for informativeness and humanitarian classification tasks, respectively. The consolidation results in a larger dataset that affords the ability to train more sophisticated models. To that end, we provide binary and multiclass classification results using CNN, FastText, and transformer based models to address informativeness and humanitarian tasks, respectively. We make the dataset and scripts available at https://crisisnlp.qcri.org/crisis_datasets_benchmarks.html.

In this paper, we study the associations between human faces and voices. Audiovisual integration, specifically the integration of facial and vocal information is a well-researched area in neuroscience. It is shown that the overlapping information between the two modalities plays a significant role in perceptual tasks such as speaker identification. Through an online study on a new dataset we created, we confirm previous findings that people can associate unseen faces with corresponding voices and vice versa with greater than chance accuracy. We computationally model the overlapping information between faces and voices and show that the learned cross-modal representation contains enough information to identify matching faces and voices with performance similar to that of humans. Our representation exhibits correlations to certain demographic attributes and features obtained from either visual or aural modality alone. We release our dataset of audiovisual recordings and demographic annotations of people reading out short text used in our studies.

People are increasingly sharing information on social media during disaster events. This information could be valuable to emergency responders, but there remain challenges for using it to inform response efforts---including filtering relevant information from the large volumes of noise. Previous research has largely focused on identifying information that can contribute to a generalized concept of situational awareness. Our work explores the value of approaching this problem from a different perspective---one of actionablity---with the idea that information relevance may vary across responder role, domain, and other factors. This approach asks how we can get the right information to the right person at the right time? We interviewed and surveyed diverse responders to understand what "actionable" information is, allowing that actionability might differ from one responder to another. Through the findings, we (a) offer a nuanced understanding of actionability and differentiate it from situational awareness; (b) describe responders' perspective of what distinguishes good information when making rapid judgments; and (c) suggest opportunities for augmenting social media use to highlight information that needs immediate attention. We offer researchers an opportunity to frame different models of actionability to suit the requirements of a responding role.

Heart valve diseases are among the leading causes of cardiac failure around the globe. Nearly 90,000 heart valve replacements occur in the USA annually. Currently, available options for heart valve replacement include bioprosthetic and mechanical valves, both of which have severe limitations. Bioprosthetic valves can last for only 10–20 years while patients with mechanical valves always require blood-thinning medications throughout the remainder of the patient’s life. Tissue engineering has emerged as a promising solution for the development of a viable, biocompatible and durable heart valve; however, a human implantable tissue engineered heart valve is yet to be achieved. In this study, a tri-leaflet heart valve structure is developed using electrospun polycaprolactone (PCL) and poly L-lactic acid (PLLA) scaffolds, and a set of in vitro testing protocol has been developed for routine manufacturing of tissue engineered heart valves. Stress-strain curves were obtained for mechanical characterization of different valves. The performances of the developed valves were hemodynamically tested using a pulse duplicator, and an echocardiography machine. Results confirmed the superiority of the PCL-PLLA heart valve compared to pure PCL or pure PLLA. The developed in vitro test protocol involving pulse duplicator and echocardiography tests have enormous potential for routine application in tissue engineering of heart valves.

Illumination is a critical element of photography and is essential for many computer vision tasks. Flash light is unique in the sense that it is a widely available tool for easily manipulating the scene illumination. We present a dataset of thousands of ambient and flash illumination pairs to enable studying flash photography and other applications that can benefit from having separate illuminations. Different than the typical use of crowdsourcing in generating computer vision datasets, we make use of the crowd to directly take the photographs that make up our dataset. As a result, our dataset covers a wide variety of scenes captured by many casual photographers. We detail the advantages and challenges of our approach to crowdsourcing as well as the computational effort to generate completely separate flash illuminations from the ambient light in an uncontrolled setup. We present a brief examination of illumination decomposition, a challenging and underconstrained problem in flash photography, to demonstrate the use of our dataset in a data-driven approach.