Natasha Jansz

Why weight? Modelling sample and observational level variability improves power in RNA-seq analyses

Variations in sample quality are frequently encountered in small RNA-sequencing experiments, and pose a major challenge in a differential expression analysis. Removal of high variation samples reduces noise, but at a cost of reducing power, thus limiting our ability to detect biologically meaningful changes. Similarly, retaining these samples in the analysis may not reveal any statistically significant changes due to the higher noise level. A compromise is to use all available data, but to down-weight the observations from more variable samples. We describe a statistical approach that facilitates this by modelling heterogeneity at both the sample and observational levels as part of the differential expression analysis. At the sample level this is achieved by fitting a log-linear variance model that includes common sample-specific or group-specific parameters that are shared between genes. The estimated sample variance factors are then converted to weights and combined with observational level weights obtained from the mean–variance relationship of the log-counts-per-million using ‘voom’. A comprehensive analysis involving both simulations and experimental RNA-sequencing data demonstrates that this strategy leads to a universally more powerful analysis and fewer false discoveries when compared to conventional approaches. This methodology has wide application and is implemented in the open-source ‘limma’ package.

Comprehensive characterization of distinct states of human naive pluripotency generated by reprogramming

Recent reports on the characteristics of naive human pluripotent stem cells (hPSCs) obtained using independent methods differ. Naive hPSCs have been mainly derived by conversion from primed hPSCs or by direct derivation from human embryos rather than by somatic cell reprogramming. To provide an unbiased molecular and functional reference, we derived genetically matched naive hPSCs by direct reprogramming of fibroblasts and by primed-to-naive conversion using different naive conditions (NHSM, RSeT, 5iLAF and t2iLGöY). Our results show that hPSCs obtained in these different conditions display a spectrum of naive characteristics. Furthermore, our characterization identifies KLF4 as sufficient for conversion of primed hPSCs into naive t2iLGöY hPSCs, underscoring the role that reprogramming factors can play for the derivation of bona fide naive hPSCs.

The Epigenetic Regulator SMCHD1 in Development and Disease

It has very recently become clear that the epigenetic modifier SMCHD1 has a role in two distinct disorders: facioscapulohumoral muscular dystrophy (FSHD) and Bosma arhinia and micropthalmia (BAMS). In the former there are heterozygous loss-of-function mutations, while both gain- and loss-of-function mutations have been proposed to underlie the latter. These findings have led to much interest in SMCHD1 and how it works at the molecular level. We summarise here current understanding of the mechanism of action of SMCHD1, its role in these diseases, and what has been learnt from study of mouse models null for Smchd1 in the decade since the discovery of SMCHD1.

Transcriptional profiling of the epigenetic regulator Smchd1.

Smchd1 is an epigenetic repressor with important functions in healthy cellular processes and disease. To elucidate its role in transcriptional regulation, we performed two independent genome-wide RNA-sequencing studies comparing wild-type and Smchd1 null samples in neural stem cells and lymphoma cell lines. Using an R-based analysis pipeline that accommodates observational and sample-specific weights in the linear modeling, we identify key genes dysregulated by Smchd1 deletion such as clustered protocadherins in the neural stem cells and imprinted genes in both experiments. Here we provide a detailed description of this analysis, from quality control to read mapping and differential expression analysis. These data sets are publicly available from the Gene Expression Omnibus database (accession numbers GSE64099 and GSE65747).

Smchd1 regulates long-range chromatin interactions on the inactive X chromosome and at Hox clusters

The regulation of higher-order chromatin structure is complex and dynamic, and a full understanding of the suite of mechanisms governing this architecture is lacking. Here, we reveal the noncanonical SMC protein Smchd1 to be a novel regulator of long-range chromatin interactions in mice, and we add Smchd1 to the canon of epigenetic proteins required for Hox-gene regulation. The effect of losing Smchd1-dependent chromatin interactions has varying outcomes that depend on chromatin context. At autosomal targets transcriptionally sensitive to Smchd1 deletion, we found increased short-range interactions and ectopic enhancer activation. In contrast, the inactive X chromosome was transcriptionally refractive to Smchd1 ablation, despite chromosome-wide increases in short-range interactions. In the inactive X, we observed spreading of trimethylated histone H3 K27 (H3K27me3) domains into regions not normally decorated by this mark. Together, these data suggest that Smchd1 is able to insulate chromatin, thereby limiting access to other chromatin-modifying proteins.In situ Hi-C and other genome-wide and imaging analyses in different mouse embryonic cell types reveal that the noncanonical SMC protein Smchd1 regulates long-range chromatin interactions and the developmental silencing of Hox genes.

Long-range chromatin interactions on the inactive X and at Hox clusters are regulated by the non-canonical SMC protein Smchd1.

The regulation of higher order chromatin structure is complex and dynamic; however we do not yet understand the full suite of mechanisms governing architecture. Here we reveal the non-canonical SMC protein Smchd1 as a novel regulator of long-range chromatin interactions, and add it to the canon of epigenetic proteins required for Hox gene regulation. The effect of losing Smchd1-dependent chromatin interactions has varying outcomes dependent on chromatin context. At autosomal targets transcriptionally sensitive to Smchd1 deletion, we find increased short-range interactions and ectopic enhancer activation. By contrast, the inactive X chromosome is transcriptionally refractive to Smchd1 ablation, despite chromosome-wide increases in short-range interactions. There we observe spreading of H3K27me3 domains into regions not normally decorated by this mark. Together these data suggest Smchd1 has the capacity to insulate the chromatin, thereby limiting access to other chromatin modifying proteins.