Spiral Genetics
Winter 2019InactiveLarge Scale Genomic Data Mining Software
Pressure-test this opportunity
Explore the risks and possibilities with a prompt for ChatGPT, Claude, or your agent.
Large Scale Genomic Data Mining Software
Explore the risks and possibilities with a prompt for ChatGPT, Claude, or your agent.
Spiral Genetics was a Seattle-based genomics infrastructure company founded in 2009 by Adina Mangubat. Over its eleven-year life, the company built software for analyzing and comparing large populations of whole human genomes — first through its Anchored Assembly algorithm for detecting structural variants, and later through the BioGraph Platform, an open-source, reference-genome-agnostic analysis system targeting national genome sequencing projects, pharma companies, and government research institutions. The company participated in Y Combinator's Winter 2019 batch, making it one of the few companies to enter YC more than a decade after founding — a signal of reset rather than early-stage momentum. [1]
Spiral Genetics failed to convert genuine technical differentiation into a sustainable business. The core tension was structural: the company released its primary product as free, open-source software at the same moment it needed to generate revenue, while simultaneously competing against well-capitalized platform incumbents — DNAnexus, Microsoft Genomics — that were absorbing the same capabilities into broader cloud offerings. A team of eight, operating on roughly $3.8M in lifetime funding, could not sustain the sales cycles required to close government and pharma contracts. [2]
The company received an SBA PPP loan of $137,300 in 2020 — retaining seven jobs — and then went quiet. [3] No formal shutdown announcement was ever published. Tracxn lists the company as "acquired" as of late 2020, though no acquirer has been publicly identified, and the designation may reflect a data artifact rather than a genuine transaction. [4] The outcome — a quiet fade rather than a clean exit — is consistent with a company that ran out of runway before its longest-cycle customers could close.
Adina Mangubat founded Spiral Genetics in Seattle in 2009 at age 22, shortly after graduating from the University of Washington with a degree in Psychology. [5] The background was unconventional for a genomics infrastructure company — most founders in the space came from computational biology, bioinformatics, or computer science. Mangubat's path into the field appears to have been driven by intellectual curiosity about the emerging possibilities of whole-genome sequencing rather than a traditional research career.
The founding team gave the company technical depth that Mangubat's own background did not provide. Becky Drees joined as Chief Science Officer and Jeremy Bruestle as Chief Technology Officer, creating a leadership core that combined scientific credibility with engineering execution. [6] The combination — a commercially-oriented CEO with a strong scientific and technical co-founding team — was a deliberate structural choice for a company selling into research institutions and government programs.
The founding insight was straightforward but well-timed: the cost of whole-genome sequencing was collapsing, and the bottleneck was shifting from data generation to data analysis. Existing bioinformatics tools were built for single-genome analysis and could not scale to the population-level comparisons that would be required as national genome sequencing programs began generating hundreds of thousands of samples. Spiral Genetics positioned itself as the infrastructure layer for that transition.
The company's early product, Anchored Assembly, focused on detecting structural variants — large-scale rearrangements in the genome that are difficult to identify with standard alignment tools. Variants of the product included Onco Assembly for oncogenic structural variants and Multi-sample Assembly for cross-sample comparison. [7] This was technically credible work in a genuinely hard problem space.
By 2019, Mangubat had articulated a sharper version of the founding thesis: "We realized was that it's not enough to analyze individual genomes quickly. What you actually want to do is you want to analyze group populations of people simultaneously." [8] That realization — that the unit of value was population-scale comparison, not individual genome analysis — drove the pivot to BioGraph and the 2019 relaunch. Whether it came early enough to matter is the central question of the company's story.
Mangubat's public profile rose quickly. She was named to Forbes 30 Under 30 in Science and Healthcare in 2013, and again as a 30 Under 30 All-Star Alumni in 2017 — recognition that reflected genuine early momentum but also created expectations the company would struggle to meet. [9]
Spiral Genetics built two distinct generations of genomics analysis software across its eleven-year life.
Generation One: Anchored Assembly (2009–2018)
The company's first product line centered on Anchored Assembly, a high-performance computational algorithm for detecting structural variants in human genomes. Structural variants — large-scale rearrangements such as deletions, duplications, inversions, and translocations — are among the most clinically significant but technically difficult features to identify in genomic data. Standard short-read sequencing tools were designed for single-nucleotide variant detection and performed poorly on structural variants at scale.
Anchored Assembly addressed this gap with two specialized variants: Onco Assembly, targeting oncogenic structural variants relevant to cancer research and diagnostics, and Multi-sample Assembly, enabling cross-sample comparison across multiple genomes simultaneously. [7] The product was sold to genetics testing companies, research institutions, and early pharma customers.
Generation Two: BioGraph Platform (2019–2020)
The 2019 relaunch introduced BioGraph — a fundamentally different architecture built around graph genomes rather than linear reference sequences. The distinction matters: traditional genomics tools align sequencing reads against a single "reference genome" (essentially an average human genome), which systematically misses variants that differ significantly from that reference. BioGraph used a graph-based representation that could encode multiple possible genome sequences simultaneously, making it reference-genome-agnostic and better suited to diverse populations. [18]
The platform was designed for three primary use cases:
The user experience was designed for bioinformaticians and computational biologists at research institutions and genome centers, not clinical end-users. The workflow involved ingesting raw whole-genome sequencing data, building or querying a graph genome index, and running population-level analyses to identify variants of interest.
BioGraph was released as free, open-source software — a deliberate choice to lower adoption barriers in a research community that was skeptical of proprietary genomics infrastructure. [18] The company filed seven patents on the underlying technology, suggesting it believed the IP was defensible even if the software itself was open. [19]
The Microsoft Healthcare partnership, announced simultaneously with the relaunch, added a specific applied use case: training machine learning algorithms on population genomic data to predict cardiovascular disease risk. [20] The partnership included financial support, Azure cloud access, and collaboration with the Microsoft Genomics team — giving BioGraph a credible cloud deployment path and a high-profile validation.
As Mangubat described the platform's core value proposition: "Existing tools are not built to handle this scale of data. We are making it possible to mine through this data to find the answers we've all been looking for." [21]
Spiral Genetics targeted three distinct customer segments at the time of its 2019 relaunch: national genome sequencing programs (government-funded initiatives to sequence large population cohorts), large academic sequencing institutions, and pharmaceutical companies conducting genomic research. [22]
At relaunch, the company had active relationships with three national genome sequencing projects, two large U.S. sequencing institutions, and was working on small projects with four countries. [14] [15] The BioGraph platform was being piloted at Baylor College of Medicine, and a collaboration with SIDRA Medical on a graph genome of the Arab population was presented at the 2019 ASHG conference. [23] [16]
These customers shared a common profile: large, technically sophisticated institutions with genuine need for population-scale genomic analysis, long procurement cycles, and limited urgency to switch from existing workflows. The customer list was impressive for a company of eight people — but the conversion from pilot to paid contract in this segment routinely takes 12–24 months, and sometimes longer for government programs.
The global genomics market was growing rapidly during Spiral Genetics' operating life, driven by the collapse in sequencing costs and the launch of national genome programs in the UK (100,000 Genomes Project), the U.S. (All of Us), Qatar, and elsewhere. The bioinformatics software segment — Spiral Genetics' specific niche — was estimated at several billion dollars annually by the late 2010s, with strong projected growth as sequencing volumes scaled.
The structural variant detection and population genomics sub-segments were smaller but strategically important: they represented the analytical bottleneck that would determine whether large sequencing investments produced clinical or research value. Spiral Genetics was correctly positioned at a real chokepoint in the genomics data pipeline.
The competitive landscape for Spiral Genetics was structurally unfavorable along the dimensions that mattered most: distribution reach and capital depth.
Platform incumbents with distribution advantages. DNAnexus had established itself as the dominant cloud genomics platform for large-scale sequencing projects, with deep relationships at the FDA, UK Biobank, and major pharma companies. Microsoft Genomics — simultaneously Spiral Genetics' partner and a structural competitor — was integrating genomics analysis directly into Azure, giving it distribution to any institution already running on Microsoft cloud infrastructure. These incumbents did not need to win on product depth; they won on the path of least resistance for procurement teams at large institutions. [24]
Open-source community alternatives. The graph genome concept was not proprietary to Spiral Genetics. The Genome Reference Consortium and academic groups were developing open-source graph genome tools (notably vg, the variation graph toolkit) that served the same research use case without licensing costs. For academic institutions with bioinformatics staff capable of running open-source tools, BioGraph's open-source release removed the one differentiator — ease of use — that might have justified a commercial relationship.
Acquirer-turned-competitor. Fabric Genomics, which had acquired and then released Spiral Genetics, remained in the competitive landscape as both a former parent and a direct competitor in genomic analysis software. [25] The relationship complicated Spiral Genetics' positioning with customers who might have had existing Fabric Genomics relationships.
The Microsoft paradox. The Microsoft Healthcare partnership was simultaneously Spiral Genetics' most valuable validation and its most dangerous competitive dynamic. Microsoft's Genomics team had overlapping technical capabilities and vastly greater resources. A partnership that began as financial support and cloud access could, over time, absorb the core technology — particularly if BioGraph's open-source codebase made the underlying algorithms available without restriction.
Spiral Genetics' strongest competitive position was on product depth: the graph genome architecture and structural variant detection capabilities were genuinely differentiated from standard linear-reference tools. But product depth is a weak competitive moat when incumbents have distribution advantages and the core technology is open-source.
Spiral Genetics never publicly disclosed its revenue model in detail, and the absence of that disclosure is itself informative. The company reported $250,000 in cumulative revenue at the time of its March 2019 relaunch — after ten years of operation. [14] That figure suggests the company had not established a repeatable commercial model during its first decade.
The intended model at relaunch appeared to be software licensing: the $1M letter of intent referenced at relaunch was described as a customer purchasing licenses. [14] Whether that LOI converted to a signed contract is unknown. The open-source release of BioGraph created a direct tension with this model: customers who could run the open-source version had no obvious reason to pay for a license.
A plausible inference is that the intended monetization was a freemium or open-core model — open-source software for research use, with paid licenses for enterprise features, support, or cloud-managed deployments. This is a common model in developer infrastructure, but it requires either a large open-source user base (to create upsell opportunities) or a clear enterprise feature set that justifies the premium. Neither condition appears to have been established before the company went quiet.
On burn rate: with a team of eight and a Seattle cost base, a rough estimate of $1.5–2M annual burn is plausible (inferring from typical salaries and overhead for a small technical team in 2019–2020). Against total lifetime funding of approximately $3.8M — with the last institutional round being the $150K YC investment in March 2019 — the company had, at best, 12–18 months of runway from the YC relaunch. [2] [13] These are inferences, not confirmed figures.
At the time of its March 2019 relaunch, Spiral Genetics had accumulated $250,000 in revenue and held a signed $1M letter of intent from a prospective customer. [14] The company was working with three national genome sequencing projects, two large U.S. sequencing institutions, and had small active projects with four countries. [14] [15]
The BioGraph platform was in active pilot at Baylor College of Medicine, and the SIDRA Medical collaboration on an Arab population graph genome was presented at the 2019 American Society of Human Genetics conference — a credible scientific venue. [23] [16] The Microsoft Healthcare partnership provided both financial support and Azure cloud access. [26]
The traction picture was genuinely encouraging for a relaunching company: multiple institutional pilots, a high-profile industry partnership, and a seven-figure LOI. The critical unknown — which no public source resolves — is whether any of these pilots converted to paying contracts after 2019. The PPP loan in 2020, retaining seven jobs, suggests the team remained intact through mid-2020 but does not indicate commercial progress. [3]
The most direct explanation for Spiral Genetics' failure is arithmetic. The company's target customers — national genome programs, government research institutions, large academic sequencing centers, and pharma companies — operate on procurement timelines that routinely run 12–24 months from initial engagement to signed contract. A pilot at Baylor College of Medicine, a collaboration with SIDRA Medical, and LOIs from national genome projects are not revenue; they are the beginning of a process that requires sustained relationship management, technical support, and organizational patience.
Spiral Genetics entered its 2019 relaunch with approximately $3.8M in lifetime funding, of which the most recent institutional contribution was $150K from YC. [2] [13] With a team of eight in Seattle, the company needed a follow-on raise to survive long enough for its pipeline to close. Mangubat expressed optimism about fundraising in May 2019 — "Fundraising can be a slog, and that's not been my experience this time around" [27] — but no follow-on round has ever been publicly announced. The PPP loan of $137,300 in 2020 was not a fundraising event; it was an emergency bridge. [3]
The attempted remedy was the YC relaunch itself — using Demo Day momentum and the Microsoft partnership announcement to catalyze a seed or Series A round. That attempt appears to have failed, or at minimum produced insufficient capital to extend the runway through contract close.
BioGraph was released as free, open-source software at the same moment the company needed to generate revenue. [18] This decision reflects a genuine strategic dilemma in developer infrastructure: proprietary software is harder to adopt in research communities that distrust vendor lock-in, but open-source software removes the most obvious monetization lever.
The open-source release was likely intended to accelerate adoption — particularly among the academic and government institutions that Spiral Genetics was targeting — with the expectation that enterprise licensing, managed cloud deployments, or support contracts would follow. This is a proven model (Red Hat, Elastic, Confluent), but it requires either a large enough user base to create upsell volume, or a clear enterprise feature tier that justifies premium pricing. Spiral Genetics had neither the scale nor the time to build either.
The company filed seven patents on the underlying technology, [19] suggesting it believed the IP was defensible even with an open-source codebase. But patents protect against copying, not against competition from well-resourced teams building similar capabilities independently. The academic graph genome community — particularly the developers of the vg toolkit — was building overlapping functionality without commercial constraints.
The attempted remedy was the $1M LOI for software licenses, suggesting the company was trying to establish a licensing model alongside the open-source release. Whether that LOI converted is unknown, and the absence of any revenue announcement after March 2019 suggests it did not close quickly.
Spiral Genetics was competing in a market where the most powerful distribution channels — cloud platforms — were simultaneously partners and structural competitors. Microsoft Genomics, operating within Azure, had the ability to offer genomic analysis capabilities to any institution already running on Microsoft cloud infrastructure, without requiring a separate procurement decision. DNAnexus had established deep relationships with the largest genome programs globally. [24]
This is a structural dynamic, not a company-specific failure. Genomics infrastructure is a category where the marginal cost of adding a new analysis capability to an existing cloud platform is low, and the switching cost for institutions already embedded in those platforms is high. A small independent vendor offering superior technical capabilities faces a fundamental distribution disadvantage: even if BioGraph was better than Microsoft Genomics' native tools, the procurement path for a large institution to adopt BioGraph required a separate vendor relationship, separate support contracts, and separate security reviews — friction that the incumbent's integrated offering eliminated.
The Microsoft partnership was an attempt to navigate this dynamic by embedding BioGraph within the Azure ecosystem rather than competing against it. But partnerships with large technology companies carry their own risks: the larger partner controls the relationship, the timeline, and ultimately the decision about whether to build the capability internally. No public evidence suggests the Microsoft partnership produced commercial revenue for Spiral Genetics.
The 2017 acquisition by Fabric Genomics and the 2019 "un-acquisition" are the most underexplored elements of Spiral Genetics' story. A company that raises a $3M Series A in 2012 and is acquired for an undisclosed amount in 2017 — after five years without a follow-on institutional round — has already exhausted one funding cycle without achieving independent scale. The acquisition was likely a soft landing, not a strategic exit.
The decision to reacquire the company and relaunch through YC in 2019 was structurally unusual. YC is designed for companies at the beginning of their journey, not for ten-year-old companies attempting a second act. The choice to use YC as a reset mechanism suggests Mangubat believed the new BioGraph product represented a genuine discontinuity from the prior company — and the Demo Day platform and investor network were the fastest path to a new funding cycle.
That framing may have been correct technically. But investors evaluating a YC W19 company with a ten-year history, a prior acquisition, and $250K in cumulative revenue were making a different risk calculation than they would for a typical seed-stage company. The "un-acquisition" narrative was compelling, but it also raised questions about why the prior institutional investors — DFJ, NuFund — were not leading a new round directly, without the YC intermediary.
The PPP loan in 2020 indicates the company was still operating through the COVID-19 period, but the pandemic likely accelerated an already-deteriorating situation. Government procurement slowed across the board in 2020. Academic institutions shifted resources to COVID-related research. The national genome programs that were Spiral Genetics' primary pipeline were disrupted. None of these factors caused the company's failure, but they compressed whatever runway remained after the post-YC fundraise fell short.
Releasing core infrastructure as open-source without a clear enterprise tier is a monetization strategy that requires scale Spiral Genetics never achieved. BioGraph was released as free, open-source software in March 2019 — the same month the company needed to raise a follow-on round. The open-source release was designed to accelerate adoption among research institutions skeptical of vendor lock-in, but it removed the most direct monetization lever at the moment the company most needed revenue. Elastic and Confluent executed similar strategies successfully, but both had tens of thousands of users before attempting to monetize. Spiral Genetics had eight employees and $250K in cumulative revenue.
Partnering with a platform incumbent that has overlapping capabilities is a distribution strategy, not a competitive moat. The Microsoft Healthcare partnership gave Spiral Genetics Azure access, financial support, and a credible co-marketing story. It did not give the company a durable competitive position. Microsoft Genomics had the resources to build graph genome capabilities internally, and any technology Spiral Genetics contributed to the partnership through its open-source codebase was available to Microsoft without restriction. Companies in developer infrastructure that partner with hyperscalers need to ensure the partnership creates dependency, not just validation.
Government and national genome program sales cycles are structurally incompatible with a small team operating on thin capital. Spiral Genetics had active pilots with three national genome sequencing projects and two large U.S. sequencing institutions at the time of its 2019 relaunch — a genuinely impressive pipeline for a team of eight. But converting those pilots to signed contracts in the government and academic research sector routinely takes 12–24 months. With approximately $3.8M in lifetime funding and no confirmed follow-on raise after March 2019, the company had, at best, 12–18 months of runway. The pipeline and the capital base were mismatched by design.
Using YC as a reset mechanism for a decade-old company creates a credibility asymmetry that is difficult to resolve in fundraising. Spiral Genetics entered YC W19 with ten years of history, a prior acquisition, and $250K in cumulative revenue. The YC brand provided Demo Day access and investor introductions, but sophisticated investors evaluating the company had to reconcile the YC framing (early-stage, high-growth potential) with the actual history (one prior institutional cycle, a soft-landing acquisition, and a relaunch). Mangubat's optimism about fundraising in May 2019 was not matched by a publicly announced round — suggesting the credibility asymmetry was harder to resolve than anticipated.
The pivot from individual genome analysis to population-scale comparison was the right insight, but it arrived after a decade of operating on the wrong unit of value. Mangubat articulated the core product pivot clearly in 2019: "It's not enough to analyze individual genomes quickly. What you actually want to do is you want to analyze group populations of people simultaneously." That insight was correct — population-scale genomics became the dominant paradigm for drug discovery and precision medicine. But Spiral Genetics spent its first decade building tools for individual genome analysis, and the pivot to BioGraph required rebuilding the product architecture from scratch. By 2019, DNAnexus and other incumbents had already established deep relationships with the national genome programs that were the natural customers for population-scale tools.