Haidong Yi

Chi-Jane Chen, Haidong Yi, Natalie Stanley

Single-cell technologies enable comprehensive profiling of diverse immune cell-types through the measurement of multiple genes or proteins per individual cell. In order to translate immune signatures assayed from blood or tissue into powerful diagnostics, machine learning approaches are often employed to compute immunological summaries or per-sample featurizations, which can be used as inputs to models for outcomes of interest. Current supervised learning approaches for computing per-sample representations are trained only to accurately predict a single outcome and do not take into account relevant additional clinical features or covariates that are likely to also be measured for each sample. Here, we introduce a novel approach for incorporating measured covariates in optimizing model parameters to ultimately specify per-sample encodings that accurately affect both immune signatures and additional clinical information. Our introduced method CytoCoSet is a set-based encoding method for learning per-sample featurizations, which formulates a loss function with an additional triplet term penalizing samples with similar covariates from having disparate embedding results in per-sample representations. Overall, incorporating clinical covariates enables the learning of encodings for each individual sample that ultimately improve prediction of clinical outcome.

Siyuan Shan, Vishal Baskaran, Haidong Yi et al.

Modern single-cell flow and mass cytometry technologies measure the expression of several proteins of the individual cells within a blood or tissue sample. Each profiled biological sample is thus represented by a set of hundreds of thousands of multidimensional cell feature vectors, which incurs a high computational cost to predict each biological sample's associated phenotype with machine learning models. Such a large set cardinality also limits the interpretability of machine learning models due to the difficulty in tracking how each individual cell influences the ultimate prediction. We propose using Kernel Mean Embedding to encode the cellular landscape of each profiled biological sample. Although our foremost goal is to make a more transparent model, we find that our method achieves comparable or better accuracies than the state-of-the-art gating-free methods through a simple linear classifier. As a result, our model contains few parameters but still performs similarly to deep learning models with millions of parameters. In contrast with deep learning approaches, the linearity and sub-selection step of our model makes it easy to interpret classification results. Analysis further shows that our method admits rich biological interpretability for linking cellular heterogeneity to clinical phenotype.

Yang Li, Haidong Yi, Christopher M. Bender et al.

Reasoning over an instance composed of a set of vectors, like a point cloud, requires that one accounts for intra-set dependent features among elements. However, since such instances are unordered, the elements' features should remain unchanged when the input's order is permuted. This property, permutation equivariance, is a challenging constraint for most neural architectures. While recent work has proposed global pooling and attention-based solutions, these may be limited in the way that intradependencies are captured in practice. In this work we propose a more general formulation to achieve permutation equivariance through ordinary differential equations (ODE). Our proposed module, Exchangeable Neural ODE (ExNODE), can be seamlessly applied for both discriminative and generative tasks. We also extend set modeling in the temporal dimension and propose a VAE based model for temporal set modeling. Extensive experiments demonstrate the efficacy of our method over strong baselines.