A deep learning–based method to infer transcription factor binding in single cells has been developed by learning from bulk data.

The advent of immune checkpoint therapy for metastatic skin cancer has greatly improved patient survival. However, most skin cancer patients are refractory to checkpoint therapy, and furthermore, the intra-immune cell signaling driving response to checkpoint therapy remains uncharacterized. When comparing the immune transcriptome in the tumor microenvironment of melanoma and basal cell carcinoma (BCC), we found that the presence of memory B cells and macrophages negatively correlate in both cancers when stratifying patients by their response, with memory B cells more present in responders. Moreover, inhibitory immune signaling mostly decreases in melanoma responders and increases in BCC responders. We further explored the relationships between macrophages, B cells and response to checkpoint therapy by developing a stochastic differential equation model which qualitatively agrees with the data analysis. Our model predicts BCC to be more refractory to checkpoint therapy than melanoma and predicts the best qualitative ratio of memory B cells and macrophages for successful treatment.

Adjuvants play a central role in enhancing the immunogenicity of otherwise poorly immunogenic vaccine antigens. Combining adjuvants has the potential to enhance vaccine immunogenicity compared with single adjuvants, although the cellular and molecular mechanisms of combination adjuvants are not well understood. Using the influenza virus hemagglutinin H5 antigen, we define the immunological landscape of combining CpG and MPLA (TLR-9 and TLR-4 agonists, respectively) with a squalene nanoemulsion (AddaVax) using immunologic and transcriptomic profiling. Mice immunized and boosted with recombinant H5 in AddaVax, CpG+MPLA, or AddaVax plus CpG+MPLA (IVAX-1) produced comparable levels of neutralizing antibodies and were equally well protected against the H5N1 challenge. However, after challenge with H5N1 virus, H5/IVAX-1–immunized mice had 100- to 300-fold lower virus lung titers than mice receiving H5 in AddaVax or CpG+MPLA separately. Consistent with enhanced viral clearance, unsupervised expression analysis of draining lymph node cells revealed the combination adjuvant IVAX-1 significantly downregulated immune homeostasis genes, and induced higher numbers of antibody-producing plasmablasts than either AddaVax or CpG+MPLA. IVAX-1 was also more effective after single-dose administration than either AddaVax or CpG+MPLA. These data reveal a novel molecular framework for understanding the mechanisms of combination adjuvants, such as IVAX-1, and highlight their potential for the development of more effective vaccines against respiratory viruses.

The advent of immune checkpoint therapy for metastatic skin cancer has greatly improved patient survival. However, most skin cancer patients are refractory to checkpoint therapy, and furthermore the differential intra-immune signaling between cancers driving the response to checkpoint therapy remains uncharacterized. When comparing the immune transcriptome in the tumor microenvironment of melanoma and basal cell carcinoma (BCC), we found that the presence of memory B cells and macrophages negatively correlate when stratifying patients by response, with memory B cells more present in responders. Inhibitory immune signaling is reduced in melanoma responders and is increased in BCC responders. We further explored the relationships between macrophages, B cells and response to checkpoint therapy by developing a stochastic differential equation model which qualitatively agrees with the data analysis. Our model predicts BCC to be more refractory to checkpoint therapy than melanoma. We show differences in tumor progression and regression that could serve as a diagnostic and predict the optimal ratio of macrophages and memory B cells for successful treatment.

Characterizing genome-wide binding profiles of transcription factor (TF) is essential for understanding many biological processes. Although techniques have been developed to assess binding profiles within a population of cells, determining binding profiles at a single cell level remains elusive. Here we report scFAN (Single Cell Factor Analysis Network), a deep learning model that predicts genome-wide TF binding profiles in individual cells. scFAN is pre-trained on genome-wide bulk ATAC-seq, DNA sequence and ChIP-seq data, and utilizes single-cell ATAC-seq to predict TF binding in individual cells. We demonstrate the efficacy of scFAN by studying sequence motifs enriched within predicted binding peaks and investigating the effectiveness of predicted TF peaks for discovering cell types. We develop a new metric “TF activity score” to characterize each cell, and show that the activity scores can reliably capture cell identities. The method allows us to discover and study cellular identities and heterogeneity based on chromatin accessibility profiles.