Marco Island, FL February 27, 2020 – AGBT’s 20th Anniversary didn’t disappoint, with many provocative sessions and product launches. In many ways, #AGBT20 mirrored the #AGBT19 vintage! This year, however, the majority of the 56 attendees we interviewed immediately highlighted spatial profiling or MGI as stealing the show. Please refer to our 2019 entry for additional commentary, as these previous trends remained relevant at this conference (e.g., biology takes center stage, NGS continued industrialization).
#1 – Spatial profiling emerges as an enabling technology (First place, 76 points, 30 mentions):
For the second year in the row, spatial profiling was the talk of the town. The three main market participants (Nanostring, 10x Genomics, Readcoor) are at various commercial stages. NanoString users displayed exciting data throughout the conference, starting with the 2nd Annual Spatial Genomics Summit on Sunday. While it is still early, we are starting to see users move from technology assessment to actual applications, most notably for the analysis of the tumor microenvironment. Interestingly, these three technologies are actually much more different in their capabilities and applications that the community currently realizes. 10x genomics offers whole transcriptome information for the entire slide that are geographically segmented in ~5,000 55 micrometer spots (typically ~10 cells, but as low as 1) – as highlighted in their Q4 2019 earnings call, the Visium has already been used in over 200 labs; NanoString offers resolution as low as 5 cells (more typically ~25) that are biologically (user) segmented (e.g., phenotypically based on low plex protein expression) currently in a few dozen regions of interest (more to come); Readcoor offers subcellular segmentation in ~100K voxels for a single cell, reading 10 – 1,000 relevant transcripts (e.g., kinases) per cells (out of the ~300,000 present) using a ~200-plex panel. We’ll have to elaborate in a white paper down the road, but the tension between plex and resolution is evident. Regardless, these platforms will all generate massive amounts of data, and data analysis / insights will most likely gate usage and adoption.
#2 – MGI is real (67 points, 31 mentions):
At the last talk of the conference, MGI announced a new platform, the DNBSEQ-Tx sequencer (“T ten”) that enables the $100 genome. While this made for a galvanizing closing talk, and generated buzz, arguably the most exciting part of the conference was that their current platforms ostensibly start to perform as advertised. While customers experienced delays on the T7 shipment, a handful of trusted conference attendees with access to a platform or the downstream data confirmed that the DNBSEQ-400 produces data “equivalent to Illumina”, and in a way that was more cost-effective. Interestingly, large tools companies (e.g., Agilent, Roche) or other providers (e.g., Sophia) are partnering with MGI, and highlighting part of their offering that is compatible with the DNBSEQ workflow. MGI appears to have a couple of additional aces up their sleeves, some of which were highlighted during the presentation (e.g., Cool-MPS). We will definitely track the speed and scale of Illumina’s competitive response.
#3 – Single Cell Analysis is commonly performed (33 points, 17 mentions):
Among lots of exciting work, we were impressed by the number of 1M cell customer experiments presented at AGBT on Chromium, particularly Sisi Chen (Caltech) and Kathryn Geiger-Schuller (Broad). The 10x Genomics Chromium Connect was on display at the conference, simplifying user workflow; the platform is now shipping. 10x Genomics also announced multiple products that will launch in 2020 (or some in 2021). For Chromium: (i) Multiomics: Single Cell ATAC + RNA (2H 2020); (ii) Multiomics: Single Cell RNA + Intracellular protein measurements with protein feature barcoding (2H 2020); (III) Targeted gene expressions with 4 fixed panels (of about 1000 genes each) + customization (mid 2020); (iv) Cell Multiplexing or CellPlex, which allows users to pool multiple samples or increase their cell loading to up to 160K recovered single cells per 10x chip (2H 2020). For Visium: (i) IHC compatibility for measuring a few protein markers at imaging resolution in conjunction with spatial gene expression (1H 2020); (ii) Targeted gene expression with 4 fixed panels (of about 1000 genes each) + customization (mid 2020); (iii) Protein feature barcoding for measurement of 10s to 100s of proteins in spatial context with NGS readout (1H 2021); (iv) FFPE compatibility and IHC kits compatible with the Visium platform (1H 2021). Many interviewees noted how quickly single cell became part of the genomics landscape, with speakers not spending any time elaborating on the methodology. Many industry stakeholders commented on the continued flawless execution from the 10x Genomics team! $250M in 2019!
#4 – Readcoor reveals the RC2, with a bang (30 points, 13 mentions):
Readcoor packed 100 people in a 500 sqft suite (with another 100 waiting outside) to reveal their instrument, the RC2! While a couple of attendees complained about the lack of available data to gauge the platform, the company certainly made a big splash! Execution is the name of the game, but they are off to a strong start.
#5 – Epigenetics excites beyond early cancer detection (18 points, 9 mentions):
David Sinclair’s talk was highlighted by most survey respondents noticing the increasing prominence of epigenetics. In our estimate, the promising field of epigenetics is about a decade beyond “standard genomics” (e.g., whole genome sequencing evaluating single nucleotide variants), with applications beyond oncology, as aging for example, as highlighted in David’s talk. The talk featured Dovetail Genomics’ Omni-C data. In a related note, Dovetail launched two new Micro-C and HiChIP kits at AGBT, expanding their suite of proximity ligation genomic solutions to assess protein-directed alterations to chromatin architecture and its resulting impact on gene expression. 10x Genomics plans to launch single cell gene expression + ATAC (epigenetics/chromatin accessibility) in the second half of 2020.
#6 – NanoString lets users do the talking (17 points, 8 mentions):
NanoString made great strides in the spatial profiling space in just a year. Joe Beechem presented impressive data collecting whole transcriptome data on the GeoMx platform. NanoString appears keenly aware of the importance of the downstream bioinformatics for data analysis, and has made progress developing (data analysis and visualization) tools to help scientists generate insights from the data.
#7 – Multiomics finds its footing (16 points, 9 mentions):
In his keynote entitled “21st Century Medicine: Systems Biology, Big Data and Deep Phenotyping “, Leroy Hood made a compelling case of the importance of analytes beyond genomics, including at single cell resolution. Of note, single cell multiomics was named the 2019 method of the year by Nature. We believe that spatial genomics will help unlock the multiomics market, and stimulate the development of a much-needed pipeline for complex data analysis.
#8 – Miroculus democratizes sample preparation expertise (12 points, 7 mentions):
Miroculus has developed a novel (oil-free electrowetting-based) digital microfluidics system to automate and miniaturize complex genomic protocols such as NGS library prep, among many other applications. The Miro Canvas can streamline lab workflow, but more disruptively accelerate protocol validation (from months in some cases to hours) by “downloading” (and uploading) protocols. For instance, a remote lab in Chile could download a Stanford protocol for RNA-Seq, and conduct it within hours. One current limitation is the lack of multiplexing, which we hope will be addressed shortly. Automation is on the rise!
#9 – TwinStrand offers near perfect accuracy sequencing (9 points, 5 mentions):
TwinStrand’s CEO Jesse Salk gave two presentations highlighting the launch of their duplex sequencing technology, an approach that can improve sequencing accuracy up to 10,000x compared to standard sequencing methods. At launch, the company will offer their platform technology for detection of minimal residual disease in acute myeloid leukemia patients (AML MRD) and genetic carcinogenic testing in preclinical testing.
#10, 11, 12 – Sequencing goes beyond standard research and clinical applications
Ancestry (9 points, 8 mentions), Forensics (9 points, 6 mentions) and Space sequencing (8 points, 6 mentions) were cited by interviewees who noticed the increasing number of NGS use cases. Shawn Levy made a compelling case of the power of genomics for forensics purposes when backed by an institution with the expertise and scale of Hudson Alpha.
#13 – Twist doubles down on NGS (8 points, 4 mentions):
Twist launched an early access program for its targeted methylation sequencing solution. In addition, the company launched a novel Comprehensive Exome Panel, an NGS target enrichment solution optimized to capture 99%+ of protein coding genes.
#14 – Long reads unmasks camouflaged genomic regions (7 points, 7 mentions):
Oxford “the great absentee” Nanopore was again multi present in talks and posters, demonstrating the value of long reads. Encouragingly, accuracy seems to continue to improve in the hands of customers but is often still (very) shy of the Q40+ accuracy that can be achieved internally by Oxford Nanopore.
#15 – Illumina is too cool for school (5 points, 5 mentions):
In a similar vein, Illumina was shockingly invisible at the conference this year, probably even more so than in 2019. This is in sharp contrast to the fact that 99%+ of the sequencing output presented at the conference likely originated from their sequencers! Remarkably, with the launch of Dragen on the NextSeq 2000, the company may have had plenty to discuss. We’ll give them the benefit of the doubt again, assuming they just want to leave room for the ecosystem to develop, and receive some credit.Many exciting products that complement the genomics ecosystem were launched at the conference. For instance, DNA Script developed an enzymatic DNA synthesizer dubbed SYNTAX; the instrument enables plates of oligos in hours ~15 minutes of hands-on time, allowing researchers to rapidly iterate and achieve same-day results. QIAGEN, Roche, and IDT also launched a battery of new products (e.g., (novel) WES products). Additional talks (Yaniv Erlich) or approaches (most notably Perturb-Seq – 4 points, 3 mentions) were mentioned by survey respondents (“Other”: 63 points, 38 mentions), continuing to demonstrate the diversity of excitement at AGBT!Finally, as usual, great additional insights can be found in Keith’s download here!
<a href='#'><img alt=' ' src='https://public.tableau.com/static/images/AG/AGBT2018TrendAnalysis/TrendAnalysis/1_rss.png' style='border: none' /></a>
Survey method: We asked interviewees what they would remember about AGBT 2020, focusing on technology, announcements, and trends. We attributed 3 points for respondent first answer, 2 points for the second answer, and 1 point for additional answers. On average, people provided ~3.5 answers.---#NGSisUnstoppable
About the Author
Stephane Budel is a partner at DeciBio with over 15 years of experience in life science business consulting, entrepreneurship and academic research. Connect with him on LinkedIn.Contact Stephane and the DeciBio team to learn the latest insights impacting your industry.Disclaimer: Companies listed above may be DeciBio clients and/or customers