Now scRNA is getting very popular, just the way next generation sequencing became very popular over the last 15 years, I was wondering, what would be the next level of sequencing?
What will be even deeper than things like scRNA, linked reads or long read sequencing?
Spatial transcriptomics is one that is already getting attention. Wider availability of direct RNA sequencing from ONT.
We are still lagging behind on actual identification of proteins in cells and metabolites. Those remain active areas of interest.
We can always use better algorithms, more efficient software for everything.
First off, transcriptomics isn't limited to what you can find from RNA-seq; there's also run-on sequencing (which hasn't really been widely applied at single-cell resolution yet).
On the subject of spatial transcriptomics, check out spatial transcriptomics at single-cell resolution to figure out where RNAs localize to within a cell. Check out: RNA seqFISH, MERFISH, APEX-Seq, etc. And of course, for those technologies as well as for other technologies (including 3D structure analysis of intracellular compartments such as chromatin), a lot of work still has yet to be done in the temporal dimension (time-course) as both ATpoint and Kevin have alluded to.
And of course, once we get good at those things, we can do perturbations (and combinatorial perturbations) to better figure out how disrupting certain genes, regulatory elements, etc. alter the transcriptome, the genome, (and the proteome) spatially and temporally.
We are making a lot of progress (considering bulk RNA-seq was first performed only around 13 years ago) but there is still so much more that can be done. The development of technologies sort of go along with trying to answer biological questions that can't be easily addressed with the current technologies in place. However, even within the scope of existing technologies, there are still so many biology questions that remain unanswered. Even within the scope of existing technologies, there is still so much work that needs to be done to improve the methods and improve the analyses of the results from those methods.
Sequencing is not everything and can only take us so far. I would appreciate the development of systems that can freeze-capture cells in a time-course, and, through this, develop 3-dimensional images of chromatin and other cellular organelles, proteins, mRNA, etc. This can likely only be be achieved via electron microscopy or nuclear magnetic resonance (NMR), but could be linked up to sequencing data, including single cell RNA-seq, single cell DNA-seq, and single cell proteomics.
With just sequencing data itself, and with extremely powerful compute resources, it should actually be possible to model chromatin formation from whole genome sequence, but this would involve input from those working in, e.g., quantum chemistry (where I briefly worked). 3C, 4C, 5C, and Hi-C, etc., cannot really do this to a great extent.
Apart from this, we have already accumulated vast amounts of information and should be doing a lot better in terms of reducing mortality from chronic diseases. In a few areas, we have seemingly learned nothing... there are certain barriers, still, toward making research truly translational. More funding for SMEs could help, as these may have fresh ideas that the larger organisations struggle to find.
Kevin
I hope it is any kind of real-time capture of biological processes. We always make snapshots of cellular states in most experiments but fail to really know where a cell comes from and where it goes to. RNA velocity is first smart approach but still in its very infancy. Processes are a continuum but we interpret it mostly as fixed states. The live tracking of chromatin remoddeling followed by the measure of RNA expression levels and subsequent detection of translation into protein would be thrilling. Live manipulation and pertubation of processes is exciting as well. Optogenetics would be a buzzword. I only heard a single talk about it so far but it was exciting even though I know too little about it to judge how promising this is in the longterm and what the limitations are.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
Indeed, ~90% of metabolites that can be identified in blood via mass spectrometry remain unidentified. We usually just filter them out for each analysis.