Biology subjects

Mace, R.

Publications and source records attributed to Mace, R..

4 recordsLinked to original sources

bioRxiv2025

Simultaneous single-cell CRISPR, RNA, and ATAC-seq enables multiomic CRISPR screens to identify gene regulatory relationships

The ability to identify gene functions and interactions in specific cellular contexts has been greatly enabled by functional genomics technologies. CRISPR-based genetic screens have proven invaluable in elucidating gene function in mammalian cells. Single-cell functional genomics methods, such as Perturb-seq and Spear-ATAC, have made it possible to achieve high-throughput mapping of the functional effects of gene perturbations by profiling transcriptomes and DNA accessibility, respectively. Combining single-cell chromatin accessibility and transcriptomic data via multiomic approaches has facilitated the discovery of novel cis and gene regulatory interactions. However, pseudobulk readouts from cell populations can often cloud the interpretation of results due to a heterogeneous response from cells receiving the same genetic perturbation, which could be mitigated by using transcriptional profiles of single cells to subset the ATAC-seq data. Existing methods to capture CRISPR guide RNAs to simultaneously assess the impact of genetic perturbations on RNA and ATAC profiles require either cloning of gRNA libraries in specialized vectors or implementing complex protocols with multiple rounds of barcoding. Here, we introduce CAT-ATAC, a technique that adds CRISPR gRNA capture to the existing 10X Genomics Multiome assay, generating paired transcriptome, chromatin accessibility and perturbation identity data from the same individual cells. We demonstrate up to 77% guide capture efficiency for both arrayed and pooled delivery of lentiviral gRNAs in induced pluripotent stem cells (iPSCs) and cancer cell lines. This capability allows us to construct gene regulatory networks (GRNs) in cells under drug and genetic perturbations. By applying CAT-ATAC, we were able to identify a GRN associated with dasatinib resistance, indirectly activated by the HIC2 gene. Using loss of function experiments, we further validated that the gene, ZFPM2, a component of the predicted GRN, also contributes to dasatinib resistance. CAT-ATAC can thus be used to generate high-content multidimensional genotype-phenotype maps to reveal novel gene and cellular interactions and functions.

genomics↗

bioRxiv2023

Kin selection underpins family dynamics in rural China

The number of co-resident members is an important characteristic of human families and thus plays an important role in human evolution. However, an inclusive fitness model of how many members should co-reside has not been explicitly tested. Here we use an evolutionary game model and a decade of family dynamic data from a rural population of southwestern China (1,110 households) to show that the existence of an evolutionary stable family size can be understood by considering the average local relatedness and average inclusive fitness of the family. Empirical data analyses show that average relatedness [Formula] decreased with increasing family size, and [Formula] was associated with helping behavior (especially for the elders) in the family, so the family tended to keep a higher level of average relatedness by adjusting the number of co-resident members. These findings are helpful to understand the variation in family size in different cultures and to demonstrate the importance of kin selection in regulating family dynamics.

evolutionary biology↗

bioRxiv2022

Optimized CRISPR guide RNA library cloning reduces skew and enables more compact genetic screens

The development of CRISPR genetic screening tools has improved functional genomics, as these tools enable precise genomic editing, provide broad access to genomic regions beyond protein-coding genes, and have fewer off-target effects than other functional genomics modalities, allowing for novel applications with smaller library sizes compared to prior technologies. Pooled functional genomics screens require high cellular coverage per perturbation to accurately quantify phenotypes and average out phenotype-independent variability across the population. While more compact libraries have decreased the number of cells needed for a given screen, the cell coverage required for large-scale CRISPR screens still poses technical hurdles to screen in more challenging systems, such as iPSC-derived and primary cells. A major factor that influences cell coverage is screening library uniformity, as larger variation in individual guide RNA abundance requires higher cell coverage to reliably measure low-abundance guides. In this work, we have systematically optimized guide RNA cloning procedures to decrease bias. We implement these protocols to demonstrate that CRISPRi screens using 10-fold fewer cells than the current standard provides equivalent statistically significant hit-calling results to screens run at higher coverage, opening the possibility of conducting genome-wide and other large-scale CRISPR screens in technically challenging models.

genomics↗

bioRxiv2017

Using Outliers in Freesurfer Segmentation Statistics to Identify Cortical Reconstruction Errors in Structural Scans

IntroductionQuality assurance (QA) is vital for ensuring the integrity of processed neuroimaging data for use in clinical neurosciences research. Manual QA (visual inspection) of processed brains for cortical surface reconstruction errors is resource-intensive, particularly with large datasets. Several semi-automated QA tools use quantitative detection of subjects for editing based on outlier brain regions. There were two project goals: (1) evaluate the adequacy of a statistical QA method relative to visual inspection, and (2) examine whether error identification and correction significantly impacts estimation of cortical parameters and established brain-behavior relationships.\n\nMethodsT1 MPRAGE images (N = 530) of healthy adults were obtained from the NKI-Rockland Sample and reconstructed using Freesurfer 5.3. Visual inspection of T1 images was conducted for: (1) participants (n = 110) with outlier values (z scores {+/-} 3 SD) for subcortical and cortical segmentation volumes (outlier group), and (2) a random sample of remaining participants (n = 110) with segmentation values that did not meet the outlier criterion (non-outlier group).\n\nResultsThe outlier group had 21% more participants with visual inspection-identified errors than participants in the non-outlier group, with a medium effect size ({Phi} = 0.22). Nevertheless, a considerable portion of images with errors of cortical extension were found in the non-outlier group (41%). Sex significantly predicted error rate; men were 2.8 times more likely to have errors than women. Although nine brain regions significantly changed size from pre- to post-editing (with effect sizes ranging from 0.26 to 0.59), editing did not substantially change the correlations of neurocognitive tasks and brain volumes (ps > 0.05).\n\nConclusionsStatistically-based QA, although less resource intensive, is not accurate enough to supplant visual inspection. We discuss practical implications of our findings to guide resource allocation decisions for image processing.

neuroscience↗