Ronin Wu

Corentin van den Broek d'Obrenan, Frédéric Galliano, Jeremy Minton et al.

The explosion of scientific publications overloads researchers with information. This is even more dramatic for interdisciplinary studies, where several fields need to be explored. A tool to help researchers overcome this is Natural Language Processing (NLP): a machine-learning (ML) technique that allows scientists to automatically synthesize information from many articles. As a practical example, we have used NLP to conduct an interdisciplinary search for compounds that could be carriers for Diffuse Interstellar Bands (DIBs), a long-standing open question in astrophysics. We have trained a NLP model on a corpus of 1.5 million cross-domain articles in open access, and fine-tuned this model with a corpus of astrophysical publications about DIBs. Our analysis points us toward several molecules, studied primarily in biology, having transitions at the wavelengths of several DIBs and composed of abundant interstellar atoms. Several of these molecules contain chromophores, small molecular groups responsible for the molecule's colour, that could be promising candidate carriers. Identifying viable carriers demonstrates the value of using NLP to tackle open scientific questions, in an interdisciplinary manner.

Jason Hoelscher-Obermaier, Edward Stevinson, Valentin Stauber et al.

The most interesting words in scientific texts will often be novel or rare. This presents a challenge for scientific word embedding models to determine quality embedding vectors for useful terms that are infrequent or newly emerging. We demonstrate how \gls{lsi} can address this problem by imputing embeddings for domain-specific words from up-to-date knowledge graphs while otherwise preserving the original word embedding model. We use the MeSH knowledge graph to impute embedding vectors for biomedical terminology without retraining and evaluate the resulting embedding model on a domain-specific word-pair similarity task. We show that LSI can produce reliable embedding vectors for rare and OOV terms in the biomedical domain.

Emily Groves, Minhong Wang, Yusuf Abdulle et al.

Automated knowledge curation for biomedical ontologies is key to ensure that they remain comprehensive, high-quality and up-to-date. In the era of foundational language models, this study compares and analyzes three NLP paradigms for curation tasks: in-context learning (ICL), fine-tuning (FT), and supervised learning (ML). Using the Chemical Entities of Biological Interest (ChEBI) database as a model ontology, three curation tasks were devised. For ICL, three prompting strategies were employed with GPT-4, GPT-3.5, BioGPT. PubmedBERT was chosen for the FT paradigm. For ML, six embedding models were utilized for training Random Forest and Long-Short Term Memory models. Five setups were designed to assess ML and FT model performance across different data availability scenarios.Datasets for curation tasks included: task 1 (620,386), task 2 (611,430), and task 3 (617,381), maintaining a 50:50 positive versus negative ratio. For ICL models, GPT-4 achieved best accuracy scores of 0.916, 0.766 and 0.874 for tasks 1-3 respectively. In a direct comparison, ML (trained on ~260,000 triples) outperformed ICL in accuracy across all tasks. (accuracy differences: +.11, +.22 and +.17). Fine-tuned PubmedBERT performed similarly to leading ML models in tasks 1 & 2 (F1 differences: -.014 and +.002), but worse in task 3 (-.048). Simulations revealed performance declines in both ML and FT models with smaller and higher imbalanced training data. where ICL (particularly GPT-4) excelled in tasks 1 & 3. GPT-4 excelled in tasks 1 and 3 with less than 6,000 triples, surpassing ML/FT. ICL underperformed ML/FT in task 2.ICL-augmented foundation models can be good assistants for knowledge curation with correct prompting, however, not making ML and FT paradigms obsolete. The latter two require task-specific data to beat ICL. In such cases, ML relies on small pretrained embeddings, minimizing computational demands.

Nouman Ahmed, Ronin Wu, Victor Botev

Finding an optimal word representation algorithm is particularly important in terms of domain specific data, as the same word can have different meanings and hence, different representations depending on the domain and context. While Generative AI and transformer architecture does a great job at generating contextualized embeddings for any given work, they are quite time and compute extensive, especially if we were to pre-train such a model from scratch. In this work, we focus on the scientific domain and finding the optimal word representation algorithm along with the tokenization method that could be used to represent words in the scientific domain. The goal of this research is two fold: 1) finding the optimal word representation and tokenization methods that can be used in downstream scientific domain NLP tasks, and 2) building a comprehensive evaluation suite that could be used to evaluate various word representation and tokenization algorithms (even as new ones are introduced) in the scientific domain. To this end, we build an evaluation suite consisting of several downstream tasks and relevant datasets for each task. Furthermore, we use the constructed evaluation suite to test various word representation and tokenization algorithms.

Adam Wesołowski, Ronin Wu, Karim Essafi

We introduce a novel class of explicit feature maps based on topological indices that represent each graph by a compact feature vector, enabling fast and interpretable graph classification. Using radial basis function kernels on these compact vectors, we define a measure of similarity between graphs. We perform evaluation on standard molecular datasets and observe that classification accuracies based on single topological-index feature vectors underperform compared to state-of-the-art substructure-based kernels. However, we achieve significantly faster Gram matrix evaluation -- up to $20\times$ faster -- compared to the Weisfeiler--Lehman subtree kernel. To enhance performance, we propose two extensions: 1) concatenating multiple topological indices into an \emph{Extended Feature Vector} (EFV), and 2) \emph{Linear Combination of Topological Kernels} (LCTK) by linearly combining Radial Basis Function kernels computed on feature vectors of individual topological graph indices. These extensions deliver up to $12\%$ percent accuracy gains across all the molecular datasets. A complexity analysis highlights the potential for exponential quantum speedup for some of the vector components. Our results indicate that LCTK and EFV offer a favourable trade-off between accuracy and efficiency, making them strong candidates for practical graph learning applications.

Aaron Mark Thomas, Yu-Cheng Chen, Hubert Okadome Valencia et al.

Navigating the vast chemical space of molecular structures to design novel drug molecules with desired target properties remains a central challenge in drug discovery. Recent advances in generative models offer promising solutions. This work presents a novel quantum circuit Born machine (QCBM)-enabled Generative Adversarial Network (GAN), called QCA-MolGAN, for generating drug-like molecules. The QCBM serves as a learnable prior distribution, which is associatively trained to define a latent space aligning with high-level features captured by the GANs discriminator. Additionally, we integrate a novel multi-agent reinforcement learning network to guide molecular generation with desired targeted properties, optimising key metrics such as quantitative estimate of drug-likeness (QED), octanol-water partition coefficient (LogP) and synthetic accessibility (SA) scores in conjunction with one another. Experimental results demonstrate that our approach enhances the property alignment of generated molecules with the multi-agent reinforcement learning agents effectively balancing chemical properties.

Mihalis Gongolidis, Jeremy Minton, Ronin Wu et al.

Domain adaptation of embedding models, updating a generic embedding to the language of a specific domain, is a proven technique for domains that have insufficient data to train an effective model from scratch. Chemistry publications is one such domain, where scientific jargon and overloaded terminology inhibit the performance of a general language model. The recent spherical embedding model (JoSE) proposed in arXiv:1911.01196 jointly learns word and document embeddings during training on the multi-dimensional unit sphere, which performs well for document classification and word correlation tasks. But, we show a non-convergence caused by global rotations during its training prevents it from domain adaptation. In this work, we develop methods to counter the global rotation of the embedding space and propose strategies to update words and documents during domain specific training. Two new document classification data-sets are collated from general and chemistry scientific journals to compare the proposed update training strategies with benchmark models. We show that our strategies are able to reduce the performance cost of domain adaptation to a level similar to Word2Vec.