STonKGs: A Sophisticated Transformer Trained on Biomedical Text and Knowledge Graphs

Abstract

The majority of biomedical knowledge is stored in structured databases or as unstructured text in scientific publications. This vast amount of information has led to numerous machine learning-based biological applications using either text through natural language processing (NLP) or structured data through knowledge graph embedding models (KGEMs). However, representations based on a single modality are inherently limited. To generate better representations of biological knowledge, we propose STonKGs, a Sophisticated Transformer trained on biomedical text and Knowledge Graphs. This multimodal Transformer uses combined input sequences of structured information from KGs and unstructured text data from biomedical literature to learn joint representations. First, we pre-trained STonKGs on a knowledge base assembled by the Integrated Network and Dynamical Reasoning Assembler (INDRA) consisting of millions of text-triple pairs extracted from biomedical literature by multiple NLP systems. Then, we benchmarked STonKGs against two baseline models trained on either one of the modalities (i.e., text or KG) across eight different classification tasks, each corresponding to a different biological application. Our results demonstrate that STonKGs outperforms both baselines, especially on the more challenging tasks with respect to the number of classes, improving upon the F1-score of the best baseline by up to 0.083. Additionally, our pre-trained model as well as the model architecture can be adapted to various other transfer learning applications. Finally, the source code and pre-trained STonKGs models are available at https://github.com/stonkgs/stonkgs and https://huggingface.co/stonkgs/stonkgs-150k.

A Systematic Literature Review on Multimodal Machine Learning: Applications, Challenges, Gaps and Future Directions

“…Identification, augmentation, screening, evaluation, measurement, and differentiation each had 1 article [162]- [167]. Language processing [168],…”

“…The use of audio modality was found in 72 articles; 5 were used individually [129], [243], [276], [284], [332], and 67 were used with others. The occupancy of text modality was found in 93 papers, and from them, it was individually used in 11 articles [48], [68], [82], [159], [168], [169], [200], [245], [316], [354], [419] and combinedly used in the rest of 82 articles. Sensor modality was identified in 40 papers, 16 of which were individually identified [65], [80], [136], [163], [166], [230]- [232], [246], [259], [270], [302], [305], [352], [394], [403] and 24 of which were grouped with others.…”

“…[95], [116], [168], [229], [244], [273], [274], [289], [300], [336], [355], [362] articles, respectively. In total, 7 and 5 articles related to similarity [48], [51], [125], [170], [171], [268], [338] and structured [64], [199], [211], [212], [262]…”