Just want to initiate a new topic discussing potential technology that might be helpful for study of metastasis. Many times the hypothesis is driven or enabled by new technology and computational analysis of technology-generated data. Number one in my mind might be single-cell RNA sequencing, which make it possible to profile those minor metastatic or stem-like clones/subpopulations in primary and metastatic tumors and DTCs floating around stroma cells or blood stream. Though there are limitations of scRNA-Seq like number of transcripts detected in each cell, it absolutely leads us one step further to study the dynamics and molecular regulations of cancer metastasis. Just FYI. My lab at St. Jude has direct and privileged access to state-of-art scRNA sequencing technology including both C1 (Fluidigm) and drop seq (10X genomics) platforms, together with powerful sequencers (>10 machines of HiSeq, NextSeq, MiSeq), in our department. My lab has generated a significant amount of scRNA-Seq data for various projects. Preliminary data of stroma (microenvironment) and tumor cells from a chemo-resistant PDX model already suggested existence of cancer stem cell subpopulations in both stroma and tumor cells that might contribute to drug resistance. Conventional or continuously-being-improved technologies like DNA-Seq, RNA-Seq and proteomics of bulk cells will likely to be needed as well for global analysis of metastasis.

Created by Jiyang Yu jiyang.yu
I believe she should be, but I just got into the office and will call in to the Webex Jiyang set up to take notes. .
@sgosline do you know if Juli is available for our meeting
use this one, please Meeting link: https://stjude.webex.com/stjude/j.php?MTID=m620d6466deb86267bdc30ae14aeeb0f1 Audio connection: 1-877-668-4490 Call-in toll-free number (US/Canada) 1-408-792-6300 Call-in toll number (US/Canada) Access code: 802 046 398
@jiyang.yu, @snyderjc1: please send webex details. thx!
Juli does not appear to be on webex @jiyang.yu i setting up his own webex meeting
@jiyang.yu and I will be discussing at 11am on Thursday if anyone would like to join.
Thanks Tony. @TonyD Yes, single-cell level network will likely correct lots of misleading connections, especially at signaling level, where we have seen many examples that the interaction is right at the bulk level but the direction is wrong. Totally agreed. Since Trey will be at the workshop, I think it's a great opportunity for us to build systematic molecular maps of the metastasis process at single-cell level using various expertise in this workshop. I would like to call a team and work on a proposal like this. Glad that @snyderjc1 has feasible mouse models that can be used to generate scRNA-Seq data. My group has access to the technology and computational tools of network inference. Anybody else who are interested are welcome to join.
@jiyang.yu perhaps more challenging to develop the systems based approach based on single-cell work, but clearly should be less susceptible to misleading network connections that might be inferred from bulk RNA-seq data...so therefore critical to do, no? I very much look forward to seeing how you develop and apply these tools in the workshop as I would expect they will be critical for developing more accurate maps of tumorigenesis and metastases...and would be very nicely paired with @snyderjc1 Crainbow mouse platform. In general, it's clearly impressive and useful to get down to single-cell resolution, but I am persuaded that getting down to just very small collections of cells (<100) is likely sufficient in which case we need better capabilities for selecting cells that belong to one population versus another before subjecting to sequencing studies or other -omic profiling. Here is where Crainbow seems to offer some really nice capabilities.
Thanks @snyderjc1. Interestingly I was about to raise another topic on models, especially in vivo or GEMMs to study metastasis. I have been searching hard on this. Will be very interested in your rainbow mouse model and how we can work together to do scRNA-Seq on your models and what questions we can answer.
I would really be interested in discussing this further. We have been trying to integrate our Cancer rainbow mouse platform with single cell seq so that we can monitor cell fate/lineage decisions and how cancer drivers can affect this in vivo. We also treat organoids with drugs (from our screening center). So we have access to great models but can't get a workflow in place for scRNAseq
Thank you for your interests and comments @TonyD. Clearly, the high discordance of mRNA and protein expression at bulk level can be partially explained by cell heterogeneity. Actually I was talking with [Junmin Peng](https://www.stjude.org/directory/p/junmin-peng.html), our proteomics director (his lab generated the best proteomics data I ever seen) about single-cell or close mass spectrum technology recently published (http://biorxiv.org/content/biorxiv/early/2017/01/23/102681.full.pdf) in parallel with scRNA-Seq we started doing a lot now, which can directly answer first your question. Long story short from the conversation with Junmin: the single-cell proteomics is not quite ready yet; the newly published technique is not really single-cell but a small amount of cells, actually Junmin's lab is optimizing a similar protocol to do that. Will be collaborating with him on some of our scRNA-Seqed samples as soon as the protocol is ready. As for your second comment, my lab is actually very active in developing computational and systems biology algorithms to infer molecular interactions or networks from scRNA-Seq data, which is clearly more challenging than network reconstruction from bulk RNA-Seq data. With lots of experience of network inference from bulk data as I did my PhD with Andrea Califano, we have made some progress on scRNA-Seq data and some predicted results are being followed-up by my collaborators. Furthermore, a senior research scientist from Mark Gerstein lab, who had a lot of network analysis experience as well, is joining my group next week and he will definitely make the progress quicker.
I am very interested in this topic and am curious if you've tried to correlate single cell protein measures with these transcripts. I am mostly thinking about the relatively high discordance between transcriptome and proteome evident from the NCI CPTAC analyses for ovarian and breast cancer samples from TCGA (BrCa 2015 - [https://www.ncbi.nlm.nih.gov/pubmed/25350482] or OvCa 2016 - [http://www.cell.com/cell/fulltext/S0092-8674(16)30673-0]). Furthermore, your experience with scRNA-Seq and where you feel the limitations for current state of the art is would be highly informative. In terms of using scRNA-seq to build maps of tumor progression and capture general heterogeneity, it would be very useful to appreciate your own perspective on how to consider reports based on this kind of data.

Technology needed: single-cell RNA-Seq and others page is loading…