For researchers across all stages of their careers, grant funding decisions can often feel like a black box. Despite this ambiguity though, one thing that holds across the proposals that succeed is the quality of the evidence they lead with. The proposals that get funded lead with data that builds the scientific case directly. The benchmarks are compelling, the evidence is concrete, and the connection between the approach and the biological question is explicit enough that the rationale for the technology choice follows from the data itself.
In our first blog post on grant writing, we covered the parts of a grant proposal that take the most time to get right, including the literature search, the competitive case, the budget, and the feasibility documentation. Now in our second installment, this post focuses on the quality of that evidence layer, specifically how to identify the benchmarks and publications that carry the most weight, and how to frame the data so it makes the scientific case on its own.
To support this in practice, we’ve built out our grant support resources, including a collection of curated papers, benchmarking data, and a free 1:1 consultation to arm you with the evidence you need to get your research funded.
Get your grant support resources
The benchmarks that carry weight
The most persuasive benchmarks are head-to-head with different sequencing approaches, run at comparable sequencing depths, and grounded in established reference datasets. Jarvis et al. (2022) is a clear example of what a compelling benchmark looks like in practice. The Human Pangenome Reference Consortium’s direct comparison of genome sequencing technologies and assembly methods for diploid human genomes found that HiFi-based assemblies produced the largest haplotype phase blocks, the lowest haplotype switch errors, the highest number of phased base pairs, and the most accurate variant calls across SNVs, small indels, and SVs in challenging genomic regions.
For indels specifically, all HiFi-based diploid assemblies achieved between 92% and 98% accuracy, compared to 52% to 58% for ONT diploid assemblies, a gap the authors attribute directly to the high indel error rate in ONT reads.
The downstream consequence is concrete. Assemblies built on ONT reads required frameshift error corrections in roughly 6,000 to 16,000 genes, compared to 100 to 200 genes for 15 kb HiF reads. Evidence of this weight doesn’t need marketing copy to support it. The data makes the case for itself.
To help you do this in your grant, the PacBio grant support resource pulls together the strongest competitive benchmarking papers and organizes them in an application- and field-specific library, so the most relevant comparisons for your proposal are findable without a separate search.
Choosing the literature to speak to your science
The most successful technology justification sections build from supporting papers where the experimental context is matched to the biology of the proposal and can be readily applied to the research question at hand.
For example, in clinical research where the sample type is highly consequential to sequencing performance, the evidence must carry across sample types. Hammond et al. (2025) illustrates this in a rigorous analytical validation of GIAB reference samples across blood, saliva, and swab samples. Their data show that HiFi WGS F1 scores for SNVs and indels exceeded 99% at approximately 15x and 25x coverage respectively and importantly, that reproducibility across two specimens and three independent sequencing datasets was above 99.8% for SNVs and above 98.6% for indels, and average high-confidence small variant concordance across specimen types was above 99.8%. For proposals in rare disease, germline variant research, or any application where specimen-type robustness is relevant, that concordance data is key.
What a strong feasibility section demonstrates
A strong feasibility section does three things. It demonstrates that the infrastructure exists or will exist, including instrument access, sample preparation capability, and compute for downstream analysis.
For a sequencing-heavy proposal, that specificity extends to instrument model and configuration, projected output per run, expected coverage at the proposed sample numbers, and the bioinformatics pipeline alongside the reference genome or assembly it will run against. The throughput math needs to close at the budget proposed, and the data volumes generated need to be compatible with the compute environment described.
These are questions most productively worked through with the technology provider directly, so a one-on-one project consultation is often particularly helpful at this stage.
What distinguishes a successful proposal
The evidence that gets a sequencing grant funded includes concrete benchmarks, matched closely to the proposed biology, and assembled thoroughly across the literature. That rigor is what separates the proposals that build genuine confidence in the work from the ones that leave questions on the table.
To get your grant support resources and schedule a 1:1 free project consultation, answer a few short questions about the grant you’re applying for and your project.