B-Cos Aligned Transformers Learn Human-Interpretable Features

Vision Transformers (ViTs) and Swin Transformers (Swin) are currently state-of-the-art in computational pathology. However, domain experts are still reluctant to use these models due to their lack of interpretability. This is not surprising, as critical decisions need to be transparent and understandable. The most common approach to understanding transformers is to visualize their attention. However, attention maps of ViTs are often fragmented, leading to unsatisfactory explanations. Here, we introduce a novel architecture called the B-cos Vision Transformer (BvT) that is designed to be more interpretable. It replaces all linear transformations with the B-cos transform to promote weight-input alignment. In a blinded study, medical experts clearly ranked BvTs above ViTs, suggesting that our network is better at capturing biomedically relevant structures. This is also true for the B-cos Swin Transformer (Bwin). Compared to the Swin Transformer, it even improves the F1-score by up to 4.7% on two public datasets.

To Transformers and Beyond: Large Language Models for the Genome

In the rapidly evolving landscape of genomics, deep learning has emerged as a useful tool for tackling complex computational challenges. This review focuses on the transformative role of Large Language Models (LLMs), which are mostly based on the transformer architecture, in genomics. Building on the foundation of traditional convolutional neural networks and recurrent neural networks, we explore both the strengths and limitations of transformers and other LLMs for genomics. Additionally, we contemplate the future of genomic modeling beyond the transformer architecture based on current trends in research. The paper aims to serve as a guide for computational biologists and computer scientists interested in LLMs for genomic data. We hope the paper can also serve as an educational introduction and discussion for biologists to a fundamental shift in how we will be analyzing genomic data in the future.

Causal machine learning for single-cell genomics

Advances in single-cell omics allow for unprecedented insights into the transcription profiles of individual cells. When combined with large-scale perturbation screens, through which specific biological mechanisms can be targeted, these technologies allow for measuring the effect of targeted perturbations on the whole transcriptome. These advances provide an opportunity to better understand the causative role of genes in complex biological processes such as gene regulation, disease progression or cellular development. However, the high-dimensional nature of the data, coupled with the intricate complexity of biological systems renders this task nontrivial. Within the machine learning community, there has been a recent increase of interest in causality, with a focus on adapting established causal techniques and algorithms to handle high-dimensional data. In this perspective, we delineate the application of these methodologies within the realm of single-cell genomics and their challenges. We first present the model that underlies most of current causal approaches to single-cell biology and discuss and challenge the assumptions it entails from the biological point of view. We then identify open problems in the application of causal approaches to single-cell data: generalising to unseen environments, learning interpretable models, and learning causal models of dynamics. For each problem, we discuss how various research directions - including the development of computational approaches and the adaptation of experimental protocols - may offer ways forward, or on the contrary pose some difficulties. With the advent of single cell atlases and increasing perturbation data, we expect causal models to become a crucial tool for informed experimental design.

Conditionally Invariant Representation Learning for Disentangling Cellular Heterogeneity

This paper presents a novel approach that leverages domain variability to learn representations that are conditionally invariant to unwanted variability or distractors. Our approach identifies both spurious and invariant latent features necessary for achieving accurate reconstruction by placing distinct conditional priors on latent features. The invariant signals are disentangled from noise by enforcing independence which facilitates the construction of an interpretable model with a causal semantic. By exploiting the interplay between data domains and labels, our method simultaneously identifies invariant features and builds invariant predictors. We apply our method to grand biological challenges, such as data integration in single-cell genomics with the aim of capturing biological variations across datasets with many samples, obtained from different conditions or multiple laboratories. Our approach allows for the incorporation of specific biological mechanisms, including gene programs, disease states, or treatment conditions into the data integration process, bridging the gap between the theoretical assumptions and real biological applications. Specifically, the proposed approach helps to disentangle biological signals from data biases that are unrelated to the target task or the causal explanation of interest. Through extensive benchmarking using large-scale human hematopoiesis and human lung cancer data, we validate the superiority of our approach over existing methods and demonstrate that it can empower deeper insights into cellular heterogeneity and the identification of disease cell states.

Training Transitive and Commutative Multimodal Transformers with LoReTTa

Collecting a multimodal dataset with two paired modalities A and B or B and C is difficult in practice. Obtaining a dataset with three aligned modalities A, B, and C is even more challenging. For example, some public medical datasets have only genetic sequences and microscopic images for one patient, and only genetic sequences and radiological images for another - but no dataset includes both microscopic and radiological images for the same patient. This makes it difficult to integrate and combine all modalities into a large pre-trained neural network. We introduce LoReTTa (Linking mOdalities with a tRansitive and commutativE pre-Training sTrAtegy) to address this understudied problem. Our self-supervised framework combines causal masked modeling with the rules of commutativity and transitivity to transition within and between different modalities. Thus, it can model the relation A -> C with A -> B -> C. Given a dataset containing only the disjoint combinations (A, B) and (B, C), we show that a transformer pre-trained with LoReTTa can handle any modality combination at inference time, including the never-seen pair (A, C) and the triplet (A, B, C). We evaluate our approach on a multimodal dataset derived from MNIST containing speech, vision, and language, as well as a real-world medical dataset containing mRNA, miRNA, and RPPA samples from TCGA. Compared to traditional pre-training methods, we observe up to a 100-point reduction in perplexity for autoregressive generation tasks and up to a 15% improvement in classification accuracy for previously unseen modality pairs during the pre-training phase.

SystemMatch: optimizing preclinical drug models to human clinical outcomes via generative latent-space matching

Translating the relevance of preclinical models ($\textit{in vitro}$, animal models, or organoids) to their relevance in humans presents an important challenge during drug development. The rising abundance of single-cell genomic data from human tumors and tissue offers a new opportunity to optimize model systems by their similarity to targeted human cell types in disease. In this work, we introduce SystemMatch to assess the fit of preclinical model systems to an $\textit{in sapiens}$ target population and to recommend experimental changes to further optimize these systems. We demonstrate this through an application to developing $\textit{in vitro}$ systems to model human tumor-derived suppressive macrophages. We show with held-out $\textit{in vivo}$ controls that our pipeline successfully ranks macrophage subpopulations by their biological similarity to the target population, and apply this analysis to rank a series of 18 $\textit{in vitro}$ macrophage systems perturbed with a variety of cytokine stimulations. We extend this analysis to predict the behavior of 66 $\textit{in silico}$ model systems generated using a perturbational autoencoder and apply a $k$-medoids approach to recommend a subset of these model systems for further experimental development in order to fully explore the space of possible perturbations. Through this use case, we demonstrate a novel approach to model system development to generate a system more similar to human biology.

Beyond Predictions in Neural ODEs: Identification and Interventions

Spurred by tremendous success in pattern matching and prediction tasks, researchers increasingly resort to machine learning to aid original scientific discovery. Given large amounts of observational data about a system, can we uncover the rules that govern its evolution? Solving this task holds the great promise of fully understanding the causal interactions and being able to make reliable predictions about the system's behavior under interventions. We take a step towards answering this question for time-series data generated from systems of ordinary differential equations (ODEs). While the governing ODEs might not be identifiable from data alone, we show that combining simple regularization schemes with flexible neural ODEs can robustly recover the dynamics and causal structures from time-series data. Our results on a variety of (non)-linear first and second order systems as well as real data validate our method. We conclude by showing that we can also make accurate predictions under interventions on variables or the system itself.

Conditional out-of-sample generation for unpaired data using trVAE

While generative models have shown great success in generating high-dimensional samples conditional on low-dimensional descriptors (learning e.g. stroke thickness in MNIST, hair color in CelebA, or speaker identity in Wavenet), their generation out-of-sample poses fundamental problems. The conditional variational autoencoder (CVAE) as a simple conditional generative model does not explicitly relate conditions during training and, hence, has no incentive of learning a compact joint distribution across conditions. We overcome this limitation by matching their distributions using maximum mean discrepancy (MMD) in the decoder layer that follows the bottleneck. This introduces a strong regularization both for reconstructing samples within the same condition and for transforming samples across conditions, resulting in much improved generalization. We refer to the architecture as \emph{transformer} VAE (trVAE). Benchmarking trVAE on high-dimensional image and tabular data, we demonstrate higher robustness and higher accuracy than existing approaches. In particular, we show qualitatively improved predictions for cellular perturbation response to treatment and disease based on high-dimensional single-cell gene expression data, by tackling previously problematic minority classes and multiple conditions. For generic tasks, we improve Pearson correlations of high-dimensional estimated means and variances with their ground truths from 0.89 to 0.97 and 0.75 to 0.87, respectively.