Concurrent Scientific Session (Genomics): Single Cell Genomics Applications
The BUS format for single-cell RNA-seq processing and analysis
The Barcode-UMI-Set format (BUS) is a recently developed format for representing pseudoalignments of reads from single-cell RNA-seq experiments. The format is can be used with most single-cell RNA-seq technologies, can be generated efficiently, and allows for development of modular and robust workflows for processing and analysis of single-cell RNA-seq reads. To demonstrate the utility of BUS, we processed 381,992,071single-cell RNA-Seq reads from a 1:1 mixture of fresh frozen human cells (HEK293T) and mouse cells (NIH3T3) produced with 10x technology and hosted on the 10x Genomics website. The generation of BUS format using a new command in the kallisto program took 984 seconds for this data (in comparison with 55,745 seconds with the 10x Genomics CellRanger software). I will present results showing that this workflow not only produces comparable results to the existing standard workflow, but is flexible and useful for many other applications.
Presenter: Fan Gao
This is joint work with Eduardo da Veiga Beltrame, Jase A. Gehring, Kristján E. Edljarn Hjoerleifsson, Lambda Lu, Páll Melsted ,Vasilis Ntranos, Lior Pachter, and Valentine Svensson.
Combination of scRNA-seq strategies to untangle complex cell populations
Single cell sequencing is changing our understanding of fundamental biological processes such as development, immunity or pathology. Current methods enable us to define the gene expression profile of thousands of cells with high resolution, allowing in depth characterization of heterogeneous cell populations such as tumours or even whole organisms. However, the integration of robust solutions for the application of these fast-evolving technologies poses a challenge to research core facilities. Moreover, the diversity of research projects and heterogeneity of samples require flexible solutions. In the Functional Genomics Center Zurich (FGCZ), we have implemented different pipelines to provide our users with access to state-of-the-art single cell technologies together with customizable analysis of single cell sequencing data. We are applying our expertise to combine different single-cell sequencing strategies adapting them to the requirements of the research and the complexity of the samples. In the case of scRNA-seq, depending on the expected heterogeneity, an initial characterization using a high throughput methodology, such as 10X Genomics, is our method of choice. This is usually the approach for tumours and clinical samples and it allows us to identify specific markers for the different cell populations. Based on the initial results or if higher resolution is needed, an automated low-volume version of the Smart-seq2 protocol is used to provide in-depth characterization of selected cell populations. The results from both approaches can be combined in our analysis pipeline, allowing us to explore highly heterogeneous samples with low proportion of relevant cells. During the talk, our experience in the implementation of a reliable and flexible single cell sequencing portfolio will be presented, illustrating how we are helping to untangle complex cell populations and to answer challenging biological questions. Moreover, an overview of recent high-throughput technologies, such as SPLIT-seq or micro-well-based solutions, and their use as standardized services offered by core facilities will be discussed.
Transcriptome profiling of cell lineage at single cell resolution with Rainbow-Seq
During pre-implantation, the totipotent 1-cell embryo undergoes cleavages to derive a blastocyst. Here, cells break symmetry during mitosis, and daughter cells diverge towards two distinct functional paths with the formation of the inner cell mass and trophectoderm. This work was motivated by our need to understand which, and how, genes are involved in the decision-making process leading to the commitment of differentiated cell types. We produced mice embryos heterozygotes for the Gt(ROSA)26Sortm1(CAG-Brainbow2.1)Cle/J construct and heterozygotes for the Tg(CAG-cre/Esr1*)5Amc/J construct. At the 2-cell stage, embryos were treated with 4-Hydroxytamoxifen in culture for one hour to promote the translocation of the CRE protein to the nuclei. CRE proteins recombined the Brainbow2.1 construct to allow for the production of one of the four possible fluorescent proteins (red, green, yellow or cyan). The embryos were then in vitro cultured and harvested at 4-, 8-cell or blastocyst stages. Blastocysts were evaluated for the binary distribution of colored cells in either the inner cell mass or trophectoderm. We identified a significant bias of sister cells towards one of the cell types in blastocysts. This result demonstrates that the cell fate could be traced back to 2-cell stage embryos. Embryos were then evaluated at 4-, 8-cell stages for the presence of fluorescence proteins; and those that showed markedly expression of one or two fluorescent proteins were split for single-cell RNA-sequencing. Using the expression of the transcript correspondent to fluorescent proteins, we assigned each blastomere to a group of cells that could be traced back to either one of the cells in the 2-cell embryo. Using these two groups we determined that few genes showed differences in expression levels that resembled the lineages. Remarkable differences were observed when we analyzed the transcriptome as a whole. With Rainbow-seq we connected transcriptome profiles with lineages at the single-cell resolution.