10X Genomics: Combining new and old techniques to unlock new insights

Illumina sequencing is by far the most common next-generation sequencing technique used today, as it extremely accurate and allows for massively parallel data generation, which is incredibly important when you’re working with a dataset the size of a human genome!

That said, there are inherent shortcomings that exist in the typical Illumina sequencing workflow. Illumina uses a very high number of short sequencing reads (usually about 150 bp for whole genome sequencing) that are then assembled together to cover the entirety of the genome. The fact that traditional Illumina can’t be used to identify long-range interactions can cause issues in some cases, such as samples with large structural variants or in phasing haplotypes.

However, a revolutionary new library preparation method designed by 10X Genomics can effectively solve these types of issues. The 10X Genomics GemCode technology is a unique reagent delivery system that allows for long-range information to be gathered from short-read sequencing. It does through usage of a high efficiency microfluidic device which releases gel beads containing unique barcodes and enzymes for library preparation. It then takes high molecular weight DNA, and partitions it into segments that are about 1M bp long; from here, these segments of DNA are combined with the gel beads. This means that each read that comes from that segment of DNA has its own unique barcode, which gives us knowledge about long-range interactions from traditional short reads.

 

 

 

10X

Figure 1: Projections of 20,000 brain cells where each cell is represented as a dot. (A) Shading highlights major clusters identified in the dataset. (B) Cells were colored based on their best match to the average expression profile of reference transcriptomes [2].

Another application of 10X Genomics GemCode comes in the form of single-cell sequencing, which also uses the microfluidics device, but combines individual cells with the gel beads instead of DNA fragments. This allows for sequencing and barcoding of individual cells from a larger heterogeneous sample. This can work for DNA or RNA sequencing. 10X Genomics recently published an application of this technology using the Chromium Single Cell 3’ Solution on a mouse brain. 1.3 million cells from embryonic mice brains were sequenced and profiled using this technique; principal component analysis and clustering was then performed on the resulting data to separate out the distinct cell types, identifying 7 major classes of cell types, as seen in Figure 1 [1].

 

Traditional Illumina will still likely reign supreme for run of the mill applications, since at this point it is still more cost effective. However, 10X is certainly gaining in popularity for specialized applications where understanding structural variants or single cell sequencing is important to the goals of the project. We’ve certainly noticed an uptick in requests for 10X genomics recently and we look forward to even more advances being made with this amazing technology.

If you’re interested in 10X technology, please contact us at projects@genohub.com for more information!

16S sequencing vs. Shotgun metagenomics: Which one to use when it comes to microbiome studies

While a lot of attention has been paid in recent years to the advances made in sequencing the human genome, next-generation sequencing has also led to an explosion of sequencing used to study microbiomes. There are two common methods of sequencing performed to study the microbiome: 16S rDNA sequencing and shotgun metagenomics.

What is 16S sequencing?

The 16S ribosomal gene is thought to exist in all bacteria, but still has regions that are highly variable between species. Because of this, primers have been created to amplify conserved regions that surround variable regions, allowing researchers to target the areas of the genes that are similar to observe the areas that are distinct. Because this approach allows us to observe very specific regions of the genome, we can drop the sequencing needed per sample dramatically, only needing around 50,000 – 100,000 reads to identify different bacterial species in a sample.

The main drawback of this technique is that it can only identify bacteria, and does not identify viruses or fungi.

What is shotgun metagenomics?

Shotgun metagenomics surveys the entire genomes of all the organisms present in the sample, as opposed to only the 16S sequences. One of the main advantages of this over 16S sequencing is that it can capture sequences from all the organisms, including viruses and fungi, which cannot be captured with 16S sequencing. Additionally, it’s less susceptible to the biases that are inherent in targeted gene amplification.

Perhaps most interestingly, it can also provide direct information about the presence or absence of specific functional pathways in sample, also known as the ‘hologenome’. This can provide potentially important information about the capabilities of the organisms in the community. Furthermore, shotgun metagenomics can be used to identify rare or novel organisms in the community, which 16S cannot do.

So which one should I use?

Like anything else, it really depends. 16S studies can be incredibly useful for comparison across different samples (like different environments, or different time points). And some studies have found that 16S sequencing is superior in these types of studies for identifying a higher number of phyla in a particular sample [1], while other studies have of course found the exact opposite [2].

When it comes down to it, it’s really important to evaluate your project needs carefully depending on what you’re trying to accomplish. For example, a large scale project that looks to examine hundreds of samples in order to evaluate the differences in microbiota across different environments might very well prefer to use 16S sequencing, since it is so much more cost-efficient than metagenomics sequencing. On the other hand, a project that is looking to deeply investigate a smaller number of samples might be a better candidate for metagenomics sequencing, which would allow them uncover all the organisms that are present in a particular sample (including viruses and fungi), as well as identify the most dominant gene pathways that are present in that particular sample.

Day 2 Summary from the Future of Genomic Medicine Conference 2018

Day 2 of the of the FOGM conference showcased some truly fascinating breakthrough talks, particularly related to how we diagnose and treat cancer.

Dr. Todd Golub of the Broad Institute gave a phenomenal talk on the need for better pre-clinical models in order to get drugs to market faster. He and his team utilized a high-throughput screening method called PRISM (Profiling Relative Inhibition Simultaneously in Mixtures), which relies on 24 nucleotide long biological barcodes that are stably integrated into genetically distinct tumor cell lines that allows screening drug candidates against a large number of cancer cell lines.

Hillary Theakston of the Clearity Foundation spoke about the importance of helping patients and families to ‘live in the future of genomic medicine’. Her foundation focuses on helping women suffering from ovarian cancer, which really lags behind in terms of survival compared to to other cancers. It’s mostly diagnosed in stage 3 or 4, and only 30% of women make it past 10 years. Clearity helps these women by spending hours of individual counseling on their unique case, including performing genetic testing of their tumor and trying to get them into drug trials that may be beneficial for their specific type of cancer–in fact, 27% of Clearity patients end up in a clinical trial. I found this talk particularly moving–while the science at the conference was incredible, it’s so important that we not forget about the patients in the process.

Dr. Mickey Kertesz, the CEO of Karius, spoke about the importance of effective clinical diagnoses. While most of the FOGM talks were cancer-related, Dr. Kertesz spoke about how we can use genomic testing to inform infectious disease diagnostics and treatments. Infectious diseases are the cause of ⅔ of the deaths of all children under 5 years old, so getting the proper treatment in a timely manner is absolutely crucial. Currently, even after 8 days in a clinical setting, only 38% of patients are diagnosed. Karius aims to improve that with their end-to-end sample processing that is able to detect circulating microbe DNA samples in the patient blood and check it for over 1,000 pathogens (possibly using 16S or metagenomic sequencing), leading to 50% of patients diagnosed in just ONE day–an enormous improvement.

Dr. C. Jimmy Lin of Natera spoke about his company’s new personalized liquid biopsy, Signatera. Signatera aims to increase the speed at which we detect cancer relapses by determining the unique clonal mutations in each patients tumor. They can then look for circulating cell-free DNA in the blood and sequence it deeply only on those specific genes (i.e., custom amplicon sequencing), looking for those same clonal mutations in the blood. Using this pipeline, they are able to detect relapses up to 100 days before they can be clinically diagnosed. The next step will be show that this can improve clinical outcomes. I wish them the best of luck with this–this could be a game-changer for diagnostics.

Dr. Kate Rubins is the first person to sequence DNA in space! It’s hard to describe how amazing this is. There are a lot of technical challenges to overcome when working in a vacuum, but Kate was able to successfully pull this off. Be able to sequence samples during space flight will certainly prove to be useful when we eventually take our first long-term mission and want to be able to sequence the samples we find ASAP!

That’s it for the Future of Genomic Medicine conference! Of course, all of the talks were fantastic and I didn’t have the space to summarize them all here. Check out the course overview here for more details.

 

Day 1 Summary from the Future of Genomic Medicine Conference 2018

 

Maker:S,Date:2017-9-5,Ver:6,Lens:Kan03,Act:Lar02,E-Y

Right next to the conference location!

It’s hard to imagine a better conference setting than the Scripps Research Institute, where you can listen to scientific talks while literally sitting right next to the beach! It’s even better when those talks are as interesting as they were. The FOGM conference covered a lot of ground:

Dr. Juan Carlos Izpisua Belmonte discussed his latest findings using the homology-independent targeted integration (or HITI) to target non-dividing cells–a key feature that means it could be used to treat adults and not just embryos. He has previously shown that they could insert treat rats with retinitis pigmentosa using this system, and they showed improvement in their ability to respond to light and healing in their eyes. In his talk at the conference, he was also able to use this same technique to alter the epigenetic landscape of mice suffering from progeria, a genetic condition that induces rapid aging, and show improved organ function and lifespan. He hopes to use this discovery to move us towards eventual treatments for the symptoms of aging–the one disease that we all suffer from. See here for the full picture.

Dr. Paul Knoepfler presented an elegant model of epigenetic effects in pediatric glioma. Pediatric gliomas are nearly 100% fatal even with the best treatments, and the treatments are incredibly severe. Children with this disease are given the same treatments as adults, but what if the tumors are different?

Well, it turns out that they are different. Dr. Knoepfler showed that pediatric gliomas frequently possess two unique point mutations in histone H3.3, and that these mutations aren’t seen in the adult gliomas. It seems astonishing that these two point mutation can convey such incredible lethality, but in fact, even small histone mutations can be incredibly lethal because of the effect on the epigenetic landscape (seen via ChIP-seq by Bjerke et al).

So, Dr. Knoepfler wanted to see if reversing these two mutations in these cancer cells could reverse the phenotype, and additionally, if doing the opposite (causing those two mutations in normal brain cells) would induce the cancer phenotype. In fact, in both cases, either reversing or causing those mutations caused an immense transcriptional shift in the opposing direction, indicating that these two point mutations are enormously important in this cancer type. Dr. Knoepfler wants to use this information to create mouse models and test new drug treatments to see which of them can be most effective against this particularly aggressive cancer.

fig

ROC curves for Dr. Mesirov’s predictive models, which outperform current clinical predictions.

Dr. Jill Mesirov also gave a very informative talk regarding pediatric brain tumors. Her lab applies machine learning and statistical techniques to find molecular markers to aid in the identification and stratification of cancer subtypes. She examined the RNA profiles from several pediatric medulloblastoma tumors using RNA-seq and found 6 different subtypes in the RNA profiles with vastly different survival rates. Using this model, she was able to categorize 15 patients that were diagnosed as ‘low-risk’ using traditional diagnostic methods as actually being ‘high-risk’–and 6 of these patients went on relapse within 3 years. Dr. Mesirov wants to use this model to help identify novel therapeutics for the particularly deadly myc-driven cancer subtype, and hopefully improve clinical outcomes.

Bonnie Rochman’s talk focused more on the sociological effects of modern genomic medicine, particularly with respects to having children. She asked some tough questions of the audience regarding prenatal testing for Down’s syndrome, or early childhood screening for genes like the BRCA1/2 genes, which predispose a person to breast cancer. Additionally, our ability to gather genomic data far outpaces our ability to accurately interpret said data, which leads to a lot of anxiety around what all this genetic testing actually means. She concludes that ultimately, there are no right or wrong answers regarding this subject–everyone has their own thoughts and feelings, shaped by their experiences with their own genetics. You can read more in her book here

Those were my favorite talks from the first day….check out the Day 2 summary tomorrow!

Day 2 Highlights from the Future of Genomic Medicine 2018 #FOGM18

The first day of the FOGM conference was absolutely incredible! I learned so much and heard some fascinating science, as well as met some truly amazing scientists and entrepreneurs. Here are the talks I’m most looking forward to today!

Todd Golub, MD, Speaking on Cancer Genomics

Dr. Golub has been a key member of important institutions such as the Broad Institute and Harvard Medical School, and made important discoveries in the genetic basis for childhood leukemia.

Mickey Kertesz, PhD, Speaking on Circulating DNA and RNA

Dr. Kertesz is the CEO of Karius, a company dedicated to making pathogen detection easier to implement for patients. He has a PhD in computational biology and did postdoctoral work at Stanford in investigating the genetic diversity of viruses.

Ash Alizadeh, MD, Phd, Speaking on Circulating DNA and RNA

Dr. Alizadeh studies the genomic biomarkers of tumors, particularly looking at non-invasive methods of detecting cancer such as looking for circulating tumor DNA (ctDNA in the blood). Shouldn’t be missed!

Jimmy Lin, MD, PhD, MHS, CSO, Speaking on Circulating DNA and RNA

Dr. Lin led the first ever exome sequencing studies in cancer and is the CSO of Natera.

Alicia Zhou, PhD, Speaking on Predictive and Preventative Genetics

Dr. Zhou studied the effects of the c-Myc oncogene in triple negative breast cancer, and currently works as the Head of Research at Color Genomics to bring population genetics to full populations.

Leslie Biesecker, MD, Speaking on Predictive and Preventative Genetics

Dr. Biesecker developed the ClinSeq® program in 2006, before the wide availability of NGS. I’m looking forward to hearing his perspective on preventative genetics.

Robert Gould, PhD, Speaking on Epigenetics

Dr. Gould has had an incredibly distinguished career. He’s currently President and CEO of Fulcrum Therapeutics–prior to that, he served as the director of novel therapeutics at the Broad Institute and spent 23 years at Merck.

Kathleen Rubins, PhD, Speaking on Our Genomics Past and Future

Dr. Rubins is the first person to sequence DNA in space! Need I say anything more?

Day 1 Highlights of the Future of Genomic Medicine Conference #FOGM18

Many of the most difficult to treat diseases that exist today have genetic origins, and one of the most difficult things about devising new treatments is the lack of connection between the research and clinical sides of biology. Because of that, the Future of Genomic Medicine conference is one of the most interesting ones to attend, because a truly fantastic mix of PhDs, MDs, and others (which this year includes journalists, CEOs and an astronaut!) have an opportunity to present and create new connections in this community.

There are so many fascinating speakers that it’s difficult to narrow it down, but here are some to watch out for on Day 1:

Eric Topol, MD, Speaking on the Future of Individualized Medicine

Dr. Topol is the founder and direction of the Scripps Translation Science Institute and in 2016 was awarded a $207M grant to lead a part of the Precision Medicine Initiative. He is one of the organizers of the Future of Genomic Medicine and has been voted the #1 influential physician leader in the US by Modern Healthcare.

Andre Choulika, PhD, Speaking on Genome Editing

In his post-doctoral work, Dr. Choulika was one of the inventors of nuclease-based genome editing and currently serves as CEO of Cellectis. We’re very interested in what he has to say on the current state of genome editing!

Paul Knoepfler, PhD, Speaking on Genome Editing

Dr. Knoepfler is not just a cancer researcher, but also a cancer survivor. He is currently studying the epigenetics of cancer and stem cells, using many techniques including CRISPR. It will be interesting to see how he uses genome editing and CRISPR in his research! He is also an active blogger and author.

Mark DePristo, PhD, Speaking on Data Science in Genomics

Dr. DePristo was part of the team that developed the GATK, one of the most prominent softwares for processing next-generation sequencing data. He currently is the head of the Genomics team at Google.

Jill P. Mesirov, PhD, Speaking on Data Science in Genomics

Dr. Mesirov does fascinating work applying machine learning to cancer genomics to stratify cancer patients according to their risk of relapse and identifying potential compounds for treatments.

Viviane Slon, Graduate Student, Speaking on Genetics of Human Origins

It’s so great to see a graduate student speaking at a conference! Viviane studies the DNA of our closest extinct relatives. It should be interesting to see her new data!

Eske Willerslev, DSc, Speaking on Genetics of Human Origins

Dr. Willerslev is an evolutionary geneticist most known for sequencing the first ancient human genome–it should be interesting to hear his perspective on human origins!

Keep an eye out for my highlights for Day 2 coming tomorrow!

 

Nanopore Sequencing: The Future of NGS?

As I mentioned in my previous post, nanopore sequencing using the MinION instrument is one of the hottest new sequencing techniques currently available. It has several benefits over the current generation of short-read sequencing instruments, including measuring epigenetic DNA modifications and ultra-long reads, which allows for improved coverage of difficult-to-sequence regions.

It does have a few drawbacks however, including the fact that it has a fairly low output, which mostly relegates it to sequencing microbial genomes. However, a recent paper by Jain et al. from UCSC [1] used the minuscule MinION instrument to sequence the human genome and compare it to the current reference genome.

There were several items of note in this paper, not the least of which is that this is the most contiguous human genome to date, getting us closer and closer to a telomere-to-telomere sequence. Additionally, they were able to close 12 gaps, each of which was more than 50 kb in length, significantly improving completion of the genome.

Amazingly, since nanopore sequencing does not utilize PCR amplification, epigenetic modifications are maintained and are actually measurable by the MinION. The instrument is capable of detecting 5-methylcytosine modifications, and this data showed good concordance with whole genome bisulfite sequencing performed in the past.

Furthermore, they were able to map several of their ultra-long reads with telomeric repeats to specific chromosomal regions. They were then able to identify the start of the telomeric repeats and calculate the length of the repeat sequence. Overall, they found evidence for repeat regions that span 2 – 11 kb.

Long and ultra-long reads are absolutely critical when it comes to annotating these highly repetitive regions. There are other sequencers, including the PacBio SMRT Sequel sequencing system, that allows for very long reads compared to the Illumina instruments. But Jain et al. were able to obtain reads that were up to a staggering 882 kb in length.

Jain et al. were able to effectively show that the MinION system is capable of being used to sequence something as complex as a human genome. Interestingly, they theorized that the MinION system may have no intrinsic limit to read length–meaning that this protocol can be improved even further by finding methods of purifying high molecular weight DNA without fragmentation. Additionally, MinION reads are still considerably less accurate than Illumina sequencing, so this aspect could be improved as well. Nonetheless, this is a truly astonishing accomplishment that indicates what the future of DNA sequencing holds in store.

If you’re interested in finding a provider of nanopore sequencing, please send us an email at projects@genohub.com and we’d love to help you with your project!