We are pleased to announce the publication of a highly improved reference genome of the mosquito Aedes aegypti, details of which were reported last week in the journal Nature. Led by Dr. Leslie Vosshall from Rockefeller University, the project to build the genome of Ae. aegypti began more than 2 years ago and involved 72 different authors from 32 institutions, including four Debug researchers from Verily. Below we discuss why this genome assembly is so important, how the broader team addressed some of the research challenges, and what Verily contributed to the effort.



Why is a good genome assembly important?

Like humans and other animals, mosquito DNA is organized into long strands called chromosomes. Not all organisms have the same number of chromosomes, for example humans have twenty-three pairs of chromosomes, while Ae. aegypti has only three. Understanding the order and sequence of DNA along the different chromosomes enables scientists to answer a wide range of questions about the biology of an organism, which in the case of the mosquito, can help us control the spread of diseases they carry.

Indeed, this new assembly of the Ae. aegypti genome has already helped researchers better understand fundamental aspects of mosquito biology including how mosquitoes find blood meals using chemosensory receptors that guide them to their human hosts, why some mosquito populations are better at transmitting disease such as dengue than others, and the molecular basis of how mosquito sex is determined.



What were some challenges?

High-quality genome assemblies have been available for insects related to Aedes, such as the fruit fly (Drosophila melanogaster) and the malaria mosquito (Anopheles gambiae), for more than a decade. Despite having roughly the same number of genes as these other species, the genome of Ae. aegypti has more than five times as much DNA! So what is all that other DNA? It is mostly repeated regions of DNA sequence known as repetitive elements, which account for more than two-thirds of the Aedes genome (Figure 1). With so much repetitive DNA, it was challenging to determine the correct order of DNA along the chromosomes, and to correctly annotate different parts of genes (exons) as they are separated by vast stretches of repeats. A previous genome assembly contained over 36,204 DNA fragments for just three chromosomes! To overcome these challenges, our team of collaborators utilized a variety of cutting-edge technologies, including a number of long-read sequencing platforms, to correctly map the genome of Ae. aegypti.



Figure 1. As shown in the pie chart above, a majority of the Aedes aegypti genome is composed of DNA repeats as opposed to unique DNA that codes for genes.



Our contribution

After a genome is correctly assembled, the next step is identifying and annotating the genes encoded. Verily produced data that helped identify the locations and sequences of genes. We did this through a process called iso-seq, which unlike many commonly employed short read technologies, allows researchers to sequence genes in their entirety. The long-read sequencing platform we used, developed by Pacific Biosciences, can achieve reads as long as 20,000 DNA base-pairs, over 60 times longer than short read techniques! Our full length transcript data helped to ensure that when scientists have a gene of interest, they will find the correct sequence in the genome.

Verily scientists also generated data highlighting the utility of the new genome. To do this, we performed whole genome sequencing on four Ae. aegypti colonies collected from different parts of the world. The sequencing data provided insight into regions of the genome that are especially diverse or conserved across the globe, an inference that would not have been possible without the new genome assembly.

Figure 2. Results of whole genome sequencing performed by the Debug team. On the left, diversity (e.g. nucleotide variation) of Aedes aegypti is visualized along the three chromosomes of the mosquito. A pronounced decrease in diversity occurs around the centromeres (center) of each chromosome. Summary statistics for each sequenced colony are given on the right.



How will the new genome assembly benefit the field going forward?

Debug is also leading an effort to sequence the genomes of 1000 wild Ae. aegypti from > 40 populations around the globe. This collaborative effort with members of the mosquito research community, will enable the Debug team to understand genetic factors that could impede efforts for mosquito control on the ground. With the improved genome assembly, we will gain insights into when this mosquito migrated out of Africa, how it has been able to adapt to so many different environments around the world, and determine the genetic flow between different populations. Such data will be critical as we scale our Wolbachia-based SIT program and bring Debug to new locations.





Sara Mitchell, PhD, Senior Scientist, Verily and Brad White, PhD, Debug Lead Scientist, Verily

As Debug Fresno’s second season comes to a close, we are excited to share our 2018 results as well as some of the key insights we have learned along the way. Around this time last year, we were wrapping up Debug Fresno’s first season and soon after, preparation for a second year began. Throughout the winter, we continued to develop our automated production and release systems and designed plans for the 2018 mosquito season with our partners, Consolidated Mosquito Abatement District (CMAD) and MosquitoMate. In April 2018, we hatched the first batch of eggs, fueled up the release vans, and embarked on a plan to conduct the world’s largest study of Wolbachia-based Sterile Insect Technique targeting Aedes aegypti.

The overall goal of our 2018 study was the same as in 2017: test whether releasing sterile male mosquitoes reared and released using Debug’s automated systems could achieve strong suppression of Aedes aegypti mosquitoes in our release areas compared to similar non-release neighborhoods. For Debug Fresno 2018, we aimed high by expanding our release areas to include three suburban Fresno County neighborhoods and increasing the variety of landscapes in the study. We also started at the beginning of the season before the wild mosquito population had the opportunity to reproduce in large numbers, in hopes that we could continue to have a large impact while releasing fewer male mosquitoes. As in 2017, we adjusted according to data and feedback from the field and compared the number of biting female Aedes aegypti in release and non-release areas at peak season to tally our final result.

The outcome was decisive. Our data indicates that we achieved a greater than 95% reduction in the number of biting, female Aedes aegypti mosquitoes compared to similar, non-release neighborhoods (confidence interval: 93.67 - 96.78). Aedes aegypti is a highly invasive mosquito that has proven difficult to combat with conventional mosquito control methods, so strong suppression across three different release areas is an important validation of a new approach to controlling these mosquitoes with a dangerous potential to transmit human disease.






Our team worked closely with our Debug Fresno partners, CMAD and MosquitoMate, to gain insights on program performance and design. Critically, our data suggest that a suppression rate beyond 95% was limited in our release sites by incursions of female Aedes aegypti mosquitoes from neighboring non-release areas, indicating that integrated control strategies and larger release areas may make this approach even more effective. We also showed that starting early in the season is a highly efficient strategy to achieve maximum suppression. This information provides invaluable guidance for future Debug programs both in the US and abroad.





As the mosquito season winds down in Fresno County and we wrap up this year’s field study, we want to sincerely thank the community residents who welcomed us in 2018. Aedes aegypti is not just a nuisance, but also a potential public health threat to the people of the Central Valley of California and actively dangerous to millions of others around the world. We are energized by what we achieved this year and look forward to bringing this technology to communities most burdened by Aedes aegypti throughout the world.





Jacob Crawford, PhD, Senior Scientist, Verily; David Clarke, PhD, Operations Manager, Verily; and Tiantian Zha, Product Manager, Verily

Today, at the Fifth Singapore International Dengue Workshop (SIDW), the National Environment Agency (NEA) announced that we have become a partner in Project Wolbachia – Singapore, a multi-year research program founded and led by Singapore’s NEA. We’re excited to bring the Debug Project’s technology to the Phase 2 field study of Project Wolbachia – Singapore, which has already shown indications that a Wolbachia-based sterile insect technique may be successful in suppressing the Aedes aegypti population in Singapore’s urban environment.






One of the unique challenges of deploying the sterile insect technique in Singapore is effectively distributing sterile, male mosquitoes across the many high-rise apartment complexes that fill its landscape. In Singapore, as in many other dense, urban environments, Aedes aegypti breed in clean, stagnant water found in planters, vases and other containers in the hallways and homes within multi-floor buildings. To tackle this challenge, we have designed a new automated release system, contained within a 1.3m x 1m cart, lightweight enough to be pushed by an individual. With this new release device, we can precisely control the distribution of sterile, male mosquitoes within the corridors of the structures. Additionally, Project Wolbachia – Singapore will employ our mosquito sex-sorting technology, which has been successfully used in Debug Fresno and Debug Innisfail to separate male and female mosquitoes using a computer vision algorithm and artificial intelligence. Our team of scientists and automation specialists look forward to collaborating with NEA to drive the science and technology of the sterile insect technique forward.

This is an exciting time for the Debug team as we enter into a new region of the world, which challenges us to consider new ecologies, mosquito behavior, and human environments. With our first field study, Debug Fresno, launching just a year and a half ago, we continue to move forward into new environments, representative of areas where dengue, Zika, and other mosquito-borne diseases commonly spread. We look forward to all we will learn in this new region and to taking another step in realizing Debug’s mission to reduce the devastating impact that disease-carrying mosquitoes inflict on people around the world.





Yanni Yoong, Program Manager, Verily and Nigel Snoad, Product Manager, Verily

During this past wet season in Far North Queensland, Debug Innisfail, a collaboration between Australia’s Commonwealth Scientific and Industrial Research Organisation (CSIRO), James Cook University (JCU) and Verily, released sterile, male Aedes aegypti mosquitoes with Wolbachia into central Innisfail and several nearby communities to research whether we could suppress the population of these dangerous invasive mosquitoes.

We came to Innisfail with CSIRO and JCU to see how Debug’s approach to reduce mosquito populations with SIT would work in a tropical environment where these mosquitoes thrive, and to learn what it was like to operate our technology with research collaborators a long way from our home base. For more than six months, our partners at CSIRO and JCU reared millions of Aedes aegypti, used our tools to remove the biting females and released more than 3 million males into three neighbourhoods around Innisfail. The CSIRO field team monitored mosquito traps and performed releases three times a week through a drought, a cyclone, and flooding in a week that brought nearly 30 inches of rain.

With the Debug Innisfail releases now finished for this season, we’re extremely happy to see the program achieve strong suppression of Aedes aegypti mosquitoes. The trap and monitoring data show that the occurrence of biting, female Aedes aegypti mosquitoes decreased by more than 80% in the three release areas relative to similar, untreated sites in the Cassowary Coast. This is a result we’re proud of and it reflects the hard work and collaboration across our partnership.

We are particularly thankful to the people of Mourilyan, South Johnstone, Goondi Bend and the rest of the Cassowary Coast region for their strong support. The community has been incredibly welcoming, hosting surveillance traps in their homes, coming to events, and supporting the team as they conducted releases.





Beyond the field results, Debug Innisfail represents an important step in the development of Verily’s Debug project: we’ve worked with and supported our partners at CSIRO and JCU as they used our technologies on the other side of the world to make and release mosquitoes. We’ve all learnt a huge amount about how to design and operate a successful Sterile Insect Technique program, and this experience is already being applied to our ongoing work.



Nigel Snoad, Product Manager, Debug

Mosquitoes in Fresno take the winter off and wait for spring. Not so for the Debug team. We spent the winter months carefully reviewing the data from our 2017 field study in Fresno, brainstorming on scientific and operational improvements, and updating our technologies and techniques. We think these updates put us in a great place to make an even bigger impact on the population of invasive Aedes aegypti in Fresno - and so on April 16th, Verily, in collaboration with Consolidated Mosquito Abatement District and MosquitoMate, began our first releases of sterile male Aedes aegypti mosquitoes of the 2018 season in Fresno County.


Like last year, the primary goal is to see a steep decline in the number of biting female Aedes aegypti mosquitoes in the areas where we release sterile male mosquitoes, but there are a few differences in the program this year. Last year, we began releasing mosquitoes in mid-July when the mosquito populations in Fresno were already approaching peak density. This year, we began releasing our sterile male mosquitoes in mid-April when mosquito reproduction is just beginning for the season, in order to outnumber and outcompete the wild male mosquito population with the release of fewer sterile male mosquitoes. In addition, we are conducting releases in more neighborhoods in Fresno County to increase the number and variety of environments included in the study.


We are pleased with what we achieved so far in Fresno County and are looking forward to continue towards the goal of reducing the number of biting, female Aedes aegypti mosquitoes during the 2018 mosquito season.


You can read more about Debug Fresno at debugfresno.com



Jacob Crawford, PhD, Senior Scientist, Verily