The situation
If you’re a postdoc or graduate student in computational biology, you’ve probably noticed it already: labs are tighter on cash. Your PI’s grant was smaller than expected. Infrastructure projects got deferred. That expensive cloud compute audit didn’t go as smoothly as it should have.
This isn’t paranoia. The National Institutes of Health (NIH), which funds roughly 44% of U.S. biomedical research, is under real budget pressure. Proposals to cap indirect costs (F&A rates) at 15% have moved from theoretical discussion to serious policy debate since late 2025. Even without a formal cap, universities are feeling the squeeze: rising operational costs, flat appropriations, and competing demands for resources.
For computational researchers, this matters more than it might for bench biologists. Cloud servers, storage, computing clusters, and software licenses come out of indirect costs. When that budget shrinks, your lab’s ability to scale experiments often shrinks with it.
The good news: you don’t have to wait for policy resolution. Here’s what’s actually happening and what you can do about it now.
What’s happening at the NIH
The backdrop is familiar. The NIH’s budget has essentially flatlined in real terms since 2009. Inflation has eaten 35% of purchasing power over that period. Meanwhile, the cost of doing research has accelerated. New technologies (sequencing, cryo-EM, cloud infrastructure) are expensive. Compliance and administrative overhead keep rising.
The indirect cost debate centers on F&A rates, which labs use to cover overhead: facility costs, library access, administrative salaries, utilities. Many elite research universities charge 50-60% of direct costs as F&A. Some proposals have suggested capping this at 15%, which would create a $30-40 billion funding shortfall for academic institutions nationally.
As of early 2026, the cap hasn’t become law. The U.S. Congress has blocked it multiple times. But the conversation itself matters: universities are already bracing for it. National Science Foundation (NSF) has quietly reduced its overhead allocations, and several major research universities have begun rewriting budget policies.
Translation: even without a formal cap, indirect funding is under pressure.
Why this hits computational biology harder
If you’re running wet-lab experiments, the budget squeeze feels like a hiring freeze or a slower grant approval timeline. You feel it, but your core work doesn’t change.
Computational biology is different. Your infrastructure costs are explicit and flexible in a way that bench work isn’t. A week of cloud compute can cost thousands. A petabyte of archival storage costs real money every year. If your lab has built a Kubernetes cluster or runs GPU jobs on AWS, that’s coming directly out of indirect costs or, worse, out of grant overhead that was supposed to fund other things.
When indirect costs shrink, labs have a choice: cut the infrastructure (which makes research slower or impossible), or eat the cost from grant money that was supposed to fund people and experiments.
This is already happening. We’re hearing from postdocs whose labs decommissioned in-house servers. We’re hearing from grad students whose co-supervisor positions got cut because cloud costs spiraled. It’s not dramatic, but it’s real.
What researchers can do right now
The key insight: don’t wait for the NIH to solve this. Diversify.
AWS Research Credits and Google Cloud for Researchers: Both AWS and Google Cloud offer substantial compute and storage credits specifically for academic researchers. These are free, separate from your grant budget, and surprisingly generous if you write a thoughtful proposal. Postdocs and grad students can apply directly (you don’t need your PI to do it). This can cover 50-80% of typical bioinformatics workloads.
NSF as a complement: The NSF is not subject to the same indirect cost pressure as NIH. If you’re doing work that touches statistics, computational methods, or fundamental tools, explore NSF grants (CAREER awards, Software Clusters, Biological Sciences programs). The review process is different from NIH, and the timeline can be faster.
Foundation funding: The Chan Zuckerberg Initiative, Gates Foundation, and other large foundations fund computational biology directly. They typically have lower overhead caps already, and they’re often faster to review than federal grants. These work best for specialized topics (single-cell methods, pandemic preparedness, disease-specific research).
Industry partnerships: This isn’t just for applied research. Many biotech and pharma companies partner with academic labs on infrastructure, sometimes supplying servers or platform access in exchange for collaboration rights. It’s worth a conversation with your PI.
Local cost optimization: Audit where your compute is really going. We’ve seen labs cut costs 30-40% just by moving batch jobs to spot instances, archiving cold data to cheaper tiers, or containerizing workflows better. This is free to do and pays immediate dividends.
A measured outlook
The NIH funding situation is real, but it’s not a sudden cliff. Policy changes, if they happen, will likely be phased or exempted for certain institution types. Universities and advocacy groups (FASEB, AAU) are fighting hard against a hard cap.
What’s certain: federal research funding will remain competitive and tight. Computational biology labs need to think of their funding like a diversified portfolio, not a single grant. That’s actually a more sustainable place to be, even if it requires more work upfront.
The researchers who will thrive over the next 3-5 years are those who learn the landscape now: who understand their local cost structure, who can articulate their computational needs precisely, and who aren’t dependent on a single funding stream.
Start with those AWS credits. Have that conversation with your PI about NSF. Look at what’s already in your institution’s contracts. Small steps compound.