The recent boom in microfluidics and combinatorial indexing strategies, combined with low sequencing costs, has empowered single-cell sequencing technology. Thousands-or even millions-of cells analyzed in a single experiment amount to a data revolution in single-cell biology and pose unique data science problems. Here, we outline eleven challenges that will be central to bringing this emerging field of single-cell data science forward. For each challenge, we highlight motivating research questions, review prior work, and formulate open problems. This compendium is for established researchers, newcomers, and students alike, highlighting interesting and rewarding problems for the coming years.

Genetic risk factors often localize to noncoding regions of the genome with unknown effects on disease etiology. Expression quantitative trait loci (eQTLs) help to explain the regulatory mechanisms underlying these genetic associations. Knowledge of the context that determines the nature and strength of eQTLs may help identify cell types relevant to pathophysiology and the regulatory networks underlying disease. Here we generated peripheral blood RNA-seq data from 2,116 unrelated individuals and systematically identified context-dependent eQTLs using a hypothesis-free strategy that does not require previous knowledge of the identity of the modifiers. Of the 23,060 significant cis-regulated genes (false discovery rate (FDR) ≤ 0.05), 2,743 (12%) showed context-dependent eQTL effects. The majority of these effects were influenced by cell type composition. A set of 145 cis-eQTLs depended on type I interferon signaling. Others were modulated by specific transcription factors binding to the eQTL SNPs.

Most disease-associated genetic variants are noncoding, making it challenging to design experiments to understand their functional consequences. Identification of expression quantitative trait loci (eQTLs) has been a powerful approach to infer the downstream effects of disease-associated variants, but most of these variants remain unexplained. The analysis of DNA methylation, a key component of the epigenome, offers highly complementary data on the regulatory potential of genomic regions. Here we show that disease-associated variants have widespread effects on DNA methylation in trans that likely reflect differential occupancy of trans binding sites by cis-regulated transcription factors. Using multiple omics data sets from 3,841 Dutch individuals, we identified 1,907 established trait-associated SNPs that affect the methylation levels of 10,141 different CpG sites in trans (false discovery rate (FDR) < 0.05). These included SNPs that affect both the expression of a nearby transcription factor (such as NFKB1, CTCF and NKX2-3) and methylation of its respective binding site across the genome. Trans methylation QTLs effectively expose the downstream effects of disease-associated variants.

Most disease associated genetic risk factors are non-coding, making it challenging to design experiments to understand their functional consequences. Identification of expression quantitative trait loci (eQTLs) has been a powerful approach to infer downstream effects of disease variants but the large majority remains unexplained.. The analysis of DNA methylation, a key component of the epigenome, offers highly complementary data on the regulatory potential of genomic regions. However, a large-scale, combined analysis of methylome and transcriptome data to infer downstream effects of disease variants is lacking. Here, we show that disease variants have wide-spread effects on DNA methylation in trans that likely reflect the downstream effects on binding sites of cis-regulated transcription factors. Using data on 3,841 Dutch samples, we detected 272,037 independent cis-meQTLs (FDR < 0.05) and identified 1,907 trait-associated SNPs that affect methylation levels of 10,141 different CpG sites in trans (FDR < 0.05), an eight-fold increase in the number of downstream effects that was known from trans-eQTL studies. Trans-meQTL CpG sites are enriched for active regulatory regions, being correlated with gene expression and overlap with Hi-C determined interchromosomal contacts. We detected many trans-meQTL SNPs that affect expression levels of nearby transcription factors (including NFKB1, CTCF and NKX2-3), while the corresponding trans-meQTL CpG sites frequently coincide with its respective binding site. Trans-meQTL mapping therefore provides a strategy for identifying and better understanding downstream functional effects of many disease-associated variants.

We show that epigenome- and transcriptome-wide association studies (EWAS and TWAS) are prone to significant inflation and bias of test statistics, an unrecognized phenomenon introducing spurious findings if left unaddressed. Neither GWAS-based methodology nor state-of-the-art confounder adjustment methods completely remove bias and inflation. We propose a Bayesian method to control bias and inflation in EWAS and TWAS based on estimation of the empirical null distribution. Using simulations and real data, we demonstrate that our method maximizes power while properly controlling the false positive rate. We illustrate the utility of our method in large-scale EWAS and TWAS meta-analyses of age and smoking.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-016-1131-9) contains supplementary material, which is available to authorized users.