The advancement of computational genomics has significantly enhanced the use of data-driven solutions in disease prediction and precision medicine. Yet, challenges such as data scarcity, privacy constraints, and biases persist. Synthetic data generation offers a promising solution to these issues. However, existing approaches based on generative artificial intelligence often fail to incorporate biological knowledge, limiting the realism and utility of generated samples. In this work, we present BioGAN, a novel generative framework that, for the first time, incorporates graph neural networks into a generative adversarial network architecture for transcriptomic data generation. By leveraging gene regulatory and co-expression networks, our model preserves biological properties in the generated transcriptomic profiles. We validate its effectiveness on E. coli and human gene expression datasets through extensive experiments using unsupervised and supervised evaluation metrics. The results demonstrate that incorporating a priori biological knowledge is an effective strategy for enhancing both the quality and utility of synthetic transcriptomic data. On human data, BioGAN achieves a 4.3% improvement in precision and an up to 2.6% higher correlation with real profiles compared to state-of-the-art models. In downstream disease and tissue classification tasks, our synthetic data improves prediction performance by an average of 5.7%. Results on E. coli further confirm BioGAN’s robustness, showing consistently strong recall and predictive utility.
Biomedical research increasingly relies on integrating diverse data modalities, including gene expression profiles, medical images, and clinical metadata. While medical images and clinical metadata are routinely collected in clinical practice, gene expression data presents unique challenges for widespread research use, mainly due to stringent privacy regulations and costly laboratory experiments. To address these limitations, we present GeMM-GAN, a novel Generative Adversarial Network conditioned on histopathology tissue slides and clinical metadata, designed to synthesize realistic gene expression profiles. GeMM-GAN combines a Transformer Encoder for image patches with a final Cross Attention mechanism between patches and text tokens, producing a conditioning vector to guide a generative model in generating biologically coherent gene expression profiles. We evaluate our approach on the TCGA dataset and demonstrate that our framework outperforms standard generative models and generates more realistic and functionally meaningful gene expression profiles, improving by more than 11\% the accuracy on downstream disease type prediction compared to current state-of-the-art generative models. Code will be available at: https://github.com/francescapia/GeMM-GAN