Evaluating Post-hoc Explanations of the Transformer-based Genome Language Model DNABERT-2
This work addresses the need for interpretable AI in genomics, enabling biological insight from transformer-based models, but is incremental as it extends existing explanation methods to a new model type.
The authors adapted AttnLRP to explain DNABERT-2, a transformer-based genome language model, and showed that its explanations reliably capture known biological patterns, similar to CNNs. They provided strategies for token- and nucleotide-level explanations and evaluated them on genomic datasets.
Explaining deep neural network predictions on genome sequences enables biological insight and hypothesis generation-often of greater interest than predictive performance alone. While explanations of convolutional neural networks (CNNs) have been shown to capture relevant patterns in genome sequences, it is unclear whether this transfers to more expressive Transformer-based genome language models (gLMs). To answer this question, we adapt AttnLRP, an extension of layer-wise relevance propagation to the attention mechanism, and apply it to the state-of-the-art gLM DNABERT-2. Thereby, we propose strategies to transfer explanations from token and nucleotide level. We evaluate the adaption of AttnLRP on genomic datasets using multiple metrics. Further, we provide an extensive comparison between the explanations of DNABERT-2 and a baseline CNN. Our results demonstrate that AttnLRP yields reliable explanations corresponding to known biological patterns. Hence, like CNNs, gLMs can also help derive biological insights. This work contributes to the explainability of gLMs and addresses the comparability of relevance attributions across different architectures.