The translational intersection of climate and biotech is underfunded relative to potential, based on an analysis of 177,962 grants made by the National Science Foundation and 188,487 publications funded by the Department of Energy.
Cancer research is mostly funded by the National Cancer Institute. Agricultural research is largely funded by the US Department of Agriculture. But does climate biotech — the deployment of bioengineering toward climate sustainability goals — have a clear funding home? It’s unclear.
Due to space constraints, we focus specifically on protein engineering, one of the fastest developing and highest potential sectors of biology. We utilized public data from the National Science Foundation and the Department of Energy to explore the funding landscape for climate biotech. All code and data is publicly available and linked throughout the article.
Our analysis suggests that only $107M from the NSF has gone to protein engineering within climate-relevant contexts since 2009. Although the Department of Energy (DoE) has a more applied mandate, fewer than 1% of its engineering-focused publications mention proteins.
This is not meant to be a critique of NSF or the DoE. We conclude that protein engineering deserves more funding attention today, but that is based on our ability to now engineer proteins in a way that would have been unimaginable in 2009. This is a privilege built on foundations established by decades of work funded by the NSF, DoE and other key funders.
Looking ahead, these results seem to match Homeworld’s hypothesis: That translationally-focused climate biotech can fall between the cracks of several government funder mandates. We present all this work to stimulate discussion and collaborate to encourage more ambitious projects towards meaningful frontiers of climate biotech.
If you are interested in this work, we encourage you to build on the Google Colabs (note that the software is “research grade”) and data files, or reach out to the Homeworld Team anytime: [email protected]
Follow along! All the code and figures for the NSF portion are available in this Google CoLab
If you go to this NSF site, you can get a list of all grants ever awarded, going back to 1959. Let’s just download all the grants since 2009 because even that is 177,962 grants! Let’s use this data to see how funding is awarded at the intersection of biology and climate.
To start analyzing the data, we load abstracts, pick a set of keywords for both climate and biology, and limit our scope to projects that have at least one climate keyword and one biology keyword.
bio_keywords = ["biology", "biochemistry","bioengineering", "biotechnology","biotech", "biochemical",
"genetics", "genomic", "omics","geobiology", "molecular biology", "cell", "biotech",
"ecology", "organism", "biodiversity", "evolution", "DNA", "RNA", "protein", "protein engineering",
"genetic engineering","enzyme", "microbiology", "immunology", "neurobiology"]
climate_keywords = ["climate change", "global warming", "greenhouse", "ozone", "deforestation",
"sustainability", "conservation", "pollution", "renewable energy", "fossil fuels",
"emissions"]
When you run the keywords filter, you get the following breakdown:
| Number of Grants | Percentage of Grants | Total Funding $ | Percentage of Funding since 2009 |
---|---|---|---|---|
Biology keywords | 67,031 | 38.1% | $38,494,731,670 | 42.9% |
Climate keywords | 23,808 | 13.5% | $13,575,899,453 | 15.1% |
Both biology and climate keywords | 11,295 | 6.4% | $7,029,467,385 | 7.8% |
Only one of the two (biology or climate) keywords | 68,249 | 38.8% | $38,011,696,353 | 42.4% |
What this says is that, since 2009, the US has invested $7B (of $94B total) into the intersection of climate and biology, which is 6.4% of the total number of projects and 7.8% of total dollars spent. Now let’s explore these 11,295 projects.
We can simply create a word cloud to visualize some statistics of the frequency of pairings between climate and biology keywords. A quick skim appears to be dominantly descriptive science:
A basic word cloud to visualize the most common biology and climate intersections
We can then dive in and quantify the top 20 pairs of biology and climate change keywords here:
The most common pairings of biology and climate keywords
And, for fun, what climate biotech pairings do we see least? We see speculative pairings such as (‘immunology’,’pollution’) and engineering projects. Overall, this is expected given these are NSF grant applications.
The least common pairings of biology and climate keywords
Some of these pairings are indeed strange. In case you are curious how the words ‘enzyme’ and ‘deforestation’ end up in the same grants, it’s a duplicated K-12 education project (no dataset is perfect) and a 2016 project (awardID 1621495) by Kevin Jarrell at Modular Genetics with the following abstract:
The broader impact/commercial potential of this Small Business Innovation Research Phase I project would be commercialization of a surfactant (acyl ethanolamine) made from renewable raw materials, which do not compete with food sources. … Successful completion of this project will demonstrate a new technology for the production of nonionic surfactants. This is significant since nonionic surfactants represent about 40% of the $30 billion surfactant market. … Bacillus strain that produces an acyl amino alcohol surfactant, namely, acyl ethanolamine. Certain naturally existing peptide synthetase enzymes catalyze the linkage of…
OK! This sounds like climate biotech to us. But, as we explore this aggregate data and dive into other specific examples, it appears that many of these projects might be educational, rather than research based. We discover via keyword matching is that 2,378 of these funded opportunities have either the word “conference” or (post-doc) “fellowship.” Let’s remove projects with these words from future analysis. Lastly, if you filter out abstracts with the word ‘education’ (not mentioned in the Broader Impacts) it shows that about 40% of the climate biotech funding was centered on training/education . In total, there are 6,798 grants since 2009 (~$3.5B total, or ~$250M/year) of funding by the National Science Foundation for projects at the intersection of biology and climate.
Now, how many of these projects are engineering-centric versus a descriptive study? As a very fast but imperfect first analysis, we can simply look for the phrase “protein engineering.” Doing a keyword match on “protein engineering” for the projects that also have climate keywords gives us only 18 hits (for a total of ~$9M), and that table of projects is listed in Appendix A. In comparison, we can check to see how many “protein engineering” keyword matches there are in the entire NSF dataset, which is about $108M, so we can conclude with a working hypothesis that 8% of protein engineering funding goes to climate efforts while presumably the majority of the other 92% likely goes to medical or agricultural applications. This assumes climate and non-climate projects use the exact phrase “protein engineering” with equal probability, and while it is a theme of many papers, only a subset of papers actually use this specific phrase. So let’s try to assess the semantics of the proposal rather than hope the authors use a new buzzword.
A more sophisticated path to assessing projects around protein engineering would be get a semantic match to the concept of protein engineering. Because this is more expensive in time and money, we start with the 6,798 research projects described above. We tried a few different approaches (eg, cosine distance between OpenAI embeddings of abstracts yielded nonsense), and ultimately the easiest for this first version of the analysis was to make a simple prompt to OpenAI’s GPT3 models.
# Running this on ~6,798 proposals cost ~$100 of OpenAI credits in August 2023
# Define the classify_abstract function as previously described
def classify_abstract(abstract):
prompt = f"This abstract describes a scientific project: '{abstract}'.
Is this an engineering project or a descriptive project? Just say 'engineering' or 'descriptive'"
response = openai.Completion.create(
engine="text-davinci-003",
prompt=prompt,
max_tokens=60
)
return response.choices[0].text.strip()
# Initialize 'project_type_guess' column with NaN values, useful if an unexpected interruption
df['project_type_guess'] = np.nan
# Apply the classify_abstract function to the 'description' column and store the result in a new column
df['project_type_guess'] = df['AbstractNarration'].progress_apply(classify_abstract)
Using OpenAI’s Completion tool yields 4,600 “Descriptive Projects” and 2,198 “Engineering Projects”. However, our choice of the keywords seemed to have added in many non-biological false positives into this analysis (eg, ‘cell’ yields work on hydrogen fuel cell which are not biological). If we limit the projects labeled as engineering to only those with ‘protein’ or ‘enzyme’ in the abstract, then there are only 211 projects.
According to this analysis, since 2009 there have been 211 protein engineering grants funded by the NSF for a total of $107M. This spreadsheet contains the entire list of these projects.
In summary, these are the high-level funding numbers for the NSF and protein engineering
Dollar amount | Description |
---|---|
$94,507,372,585 | Total sum of all NSF Grants given since 2009 |
$107,570,653 | Sum of NSF awards since 2009 for protein engineering applied to climate, as determined by semantic match using LLM tools. |
8% | Of protein engineering work funded by the NSF is in the context of climate, as determined by keyword match. |
0.1% | Percentage of protein engineering efforts in climate biotech as the overall NSF budget |
Follow along! All the code and figures for the DoE portion are available in this google CoLab
NSF has the mandate of fundamental science and education, but what about the areas of government funding which are specifically translational? For this we can turn to OSTI.gov, U.S. Department of Energy’s Office of Scientific and Technical Information.
Using the OSTI API, we downloaded all 188,487 papers as of June 19, 2023 that are publications coming from DoE funding. Note that in that the NSF analysis we were studying grants, and in this DOE analysis we’re studying publications. This means we can also utilize the CrossRef API to download all metadata for the publications too, which enables future work through metadata.
First, let’s just focus on any projects which have ‘protein’ or ‘enzyme’ in them, which immediately drops 188,487 publications to 9,125. Of these entries, we drop about ~600 that have incomplete records (again, no dataset is perfect), giving us a total working set of papers of size 8,539. From here, we can ask which institutions are doing the most published work using Department of Energy funding.
If we do a keyword match, the phrase ‘protein engineering’ drops the count from 8,539 to 125 papers. You can see where this work was done in Figure 4, and note that, for clarity, we omit any institution with “National Lab” in the title. Also keep in mind that each publication may have multiple institutions involved, which is why the bar chart sum will be great than 125.
The 30 most common institutions that published papers with the words “protein” or “enzyme” using Department of Energy funding. This list intentionally removed institutions with the phrase “National Lab” in the title as National Labs would dominate this list.
Here is a spreadsheet if you’d like to see what every single one of these 125 protein engineering papers are. Across these 125 projects are 298 climate biotech funders who fund protein engineering projects. We explored a few visualization of the co-funders (see Figure 5), demonstrated below, but decided not to pursue mapping of funders further: We found it most useful just to get a sense of how many different funding agencies there across just 125 publications.
A circular dendrogram visualization of 298 funders referenced from 125 DoE-funded publications. This was not necessarily useful but did make for a nice graphic
However, 125/8,539 projects feels like an underestimate and there is a clear reason: Just like with the NSF analysis, very few protein engineering papers explicitly use the phrase “protein engineering.” Again, we use the powerful LLMs to ask an AI to do the classification for us across all DoE projects that mention proteins. We use the following code block to ask GPT3.5 whether it would classify an abstract as ‘descriptive’ or ‘engineering.’ This cost about $60 in OpenAI API usage, and the results are saved and available via the CoLab.
def classify_abstract(abstract):
prompt = f"""This abstract describes a scientific project: '{abstract}'.
Is this an engineering project or a descriptive project?
Just say 'engineering' or 'descriptive'"""
response = openai.Completion.create(
engine="text-davinci-003",
prompt=prompt,
max_tokens=60
)
return response.choices[0].text.strip()
# Apply the classify_abstract function to the 'description' column and store the result in a new column
df_protein['project_type_guess'] = df_protein['description'].progress_apply(classify_abstract)
The result of this approach is that we get 2,302 of 8,539 papers are protein engineering efforts. If you want to see all the raw data behind the 2303 papers, you can see the spreadsheet here. As we noticed occasional biomedical papers, we can use a similar prompt to flag medical papers while also creating two keywords per abstract. You can see a set of sample results of the code block below in Appendix B, which made us confident that the LLM was doing a quite decent job.
# Running this on ~2302 proposals cost ~$20 of OpenAI credits in August 2023
# Define the classify_abstract function as previously described
def classify_translationarea(abstract):
prompt = f"""This abstract describes a scientific project: '{abstract}'.
Can you please give me one or two keywords for the target translation area
in the format of (keyword 1, keyword 2)? Additionally, please say 'medical'
or 'non-medical' if the application area is medically related.
Examples could be
(agriculture, drought adaptation, non-medical) or
(metal mining, industrial catalysis, non-medical)
(protein binding, covid19, medical)
"""
response = openai.Completion.create(
engine="text-davinci-003",
prompt=prompt,
max_tokens=60
)
return response.choices[0].text.strip()
# Initialize 'translation_area_guess' column with NaN values
df_engineering_projects['translation_area_guess'] = np.nan
# Apply the classify_abstract function to the 'description' column and store the result in a new column
df_engineering_projects['translation_area_guess'] = df_engineering_projects['description'].progress_apply(classify_translationarea)
From this we learn that there are 1,861 protein engineering papers that focus on non-medical biotech. You can view all of these papers in this spreadsheet, play with a dynamic UMAP in the Google CoLab link (using the OpenAI embedding then HDBSCAN-based clustering), or simply see a static version of the UMAP below in Figure 6. For a table of keywords from each cluster, you can refer to Appendix C.
A UMAP of the DoE publication abstracts (using the OpenAI embeddings of the abstracts) for non-medical protein engineering.
In summary, these are the high-level numbers for the Department of Energy’s funding of protein engineering projects:
Number of Papers | Description |
---|---|
188,487 | Papers published with DoE funding |
9,125 | Papers published with DoE funding with “protein” or “enzyme” in the abstract |
1,861 | Protein engineering papers published with DoE funding in a non-medical context |
< 1% | Percentage DoE publications focusing on climate-relevant protein engineering. |
The high-level takeaways are:
Protein engineering outside of medicine is deeply underfunded. $107M from the NSF since 2009 is not much: this is equivalent to a single Series A investment of a contemporary biotech company. Especially as proteins may be one day be considered as programmable nanomachines, it seems like an underinvestment of the DoE that <1% of their funding goes to a foundational new component of biotech.
Who are these people doing the work today, what are they working on? Can we create a semantic map or graph of this research? For example, where are the great protein studies by domain enzymologists that could one day collaborate with protein engineers? Answering these questions will be for another article!
There must be a focused funding vehicle to support ambitious protein engineering efforts toward climate goals.
Thank you for reading! If you have any questions, comments or suggestions to improve this analysis, please feel free to reach out to [email protected].
AwardTitle | StartDate | Amount | Organization Name |
---|---|---|---|
EAGER: Rational Modification of Enzyme Charge for Enhanced Biocatalyst Stability in Ionic Liquids | 07/24/2013 | $84,487 | REGENTS OF THE UNIVERSITY OF COLORADO, THE |
STTR Phase I: Engineering Polysaccharide Monooxygenases for Enhanced Sugar Recovery From Biomass | 07/01/2013 | $225,000 | PROTABIT LLC |
STTR Phase I: Engineering a recombinant methane monooxygenase to convert methane to methanol for the production of fuels and chemicals | 12/11/2013 | $225,000 | PROTABIT LLC |
SusChEM: Novel 1,2-Propanediol Biosynthesis from Renewable Feedstocks through Enzyme Discovery | 07/25/2014 | $317,611 | BOARD OF REGENTS OF THE UNIVERSITY OF NEBRASKA |
Synthesis and Directed Assembly of Bio-Hybrid Materials with Membrane-Protein-Mediated Transport Performance | 07/12/2014 | $420,000 | TRUSTEES OF THE COLORADO SCHOOL OF MINES |
FMRG: Bio: CAS: Distributed methane conversion into value chemicals via synthetic microbial consortia | 08/30/2022 | $3,170,876 | UNIVERSITY OF ARIZONA |
Enhancing cellulase activity through single-molecule imaging and protein engineering as a testbed for understanding and improving enzymatic deconstruction of insoluble substrates | 04/11/2023 | $629,976 | PENNSYLVANIA STATE UNIVERSITY, THE |
STTR Phase II: Development of a computational protein engineering platform and its application to methane activating enzymes | 09/14/2015 | $750,000 | PROTABIT LLC |
CAREER: Mechanism of Cytochrome P450 Alkene Biosynthesis | 02/12/2016 | $695,875 | UNIVERSITY OF SOUTH CAROLINA |
SusChEM: Engineering E. coli for improved production of polyhydroxyalkanoate (PHA)-based biodegradable plastics | 06/05/2013 | $500,000 | THE RESEARCH FOUNDATION FOR THE STATE UNIVERSITY OF NEW YORK |
Collaborative Research: A Systems Biology Approach for Metabolically Engineering Escherichia coli for Producing Hydrogen via Fermentation | 12/15/2011 | $59,869 | PENNSYLVANIA STATE UNIVERSITY, THE |
UC Davis ChemEnergy REU Site: Chemistry Research Experience for Undergraduates in Energy and Catalysis | 08/18/2010 | $335,850 | UNIVERSITY OF CALIFORNIA, DAVIS |
CAREER: Mechanism of Cytochrome P450 Alkene Biosynthesis | 11/13/2020 | $55,847 | NORTH CAROLINA STATE UNIVERSITY |
CAS:Scalable platform for materials fabrication from genetically engineered bacterial biomass | 06/15/2020 | $458,276 | NORTHEASTERN UNIVERSITY |
SBIR Phase I: Microbial conversion of pectin-rich agricultural waste into specialty chemicals | 06/13/2018 | $225,000 | ZESTBIO, INC. |
Development of Bio-Inspired Synthetic Metallopeptidases: Insight from Theoretical Studies | 06/27/2012 | $351,000 | UNIVERSITY OF MIAMI |
Synthesis and Directed Assembly of Bio-Hybrid Materials with Membrane-Protein-Mediated Transport Performance | 01/04/2016 | $398,321 | TEXAS TECH UNIVERSITY HEALTH SCIENCES CENTER |
description | translation_area_guess |
---|---|
The vast size of chemical space necessitates computational approaches to automate and accelerate the design of molecular sequences to guide experimental efforts for drug discovery. Genetic algorithms provide a useful framework to incrementally generate molecules by applying mutations to known chemical structures. Recently, masked language models have been applied to automate the mutation process by leveraging large compound libraries to learn commonly occurring chemical sequences (i.e., using tokenization) and predict rearrangements (i.e., using mask prediction). Here, we consider how language models can be adapted to improve molecule generation for different optimization tasks. We use two different generation strategies for comparison, fixed and adaptive. …. | (chemical design, language modeling, non-medical) |
(Rigid Body Docking, Sequence-Independent, Non-Medical) | |
Siderophores belonging to the ferrichrome family are essential for the viability of fungal species and play a key role for virulence of numerous pathogenic fungi. Despite their biological significance, our understanding of how these iron-chelating cyclic hexapeptides are assembled by non-ribosomal peptide synthetase (NRPS) enzymes remains poorly understood, primarily due to the nonlinearity exhibited by the domain architecture. Herein, we report the biochemical characterization of the SidC NRPS … | (non-ribosomal peptide synthetase, siderophore, non-medical) |
The state-of-art protein structure prediction methods such as AlphaFold are being widely used to predict structures of uncharacterized proteins in biomedical research. There is a significant need to further improve the quality and nativeness of the predicted structures to enhance their usability. In this work, we develop ATOMRefine, a deep learning-based, end-to-end, all-atom protein structural model refinement method. … | (protein structure, deep learning, non-medical) |
Emerging and re-emerging viral pathogens present a unique challenge for anti-viral therapeutic development. Anti-viral approaches with high flexibility and rapid production times are essential for combating these high-pandemic risk viruses. CRISPR-Cas technologies have been extensively repurposed to treat a variety of diseases, with recent work expanding into potential applications against viral infections. However, delivery still presents a major challenge for these technologies. Lipid-coated mesoporous silica nanoparticles (LCMSNs) offer an attractive delivery vehicle … | (CRISPR-Cas9, viral infection, medical) |
Microalgal biomass is poised to become an important feedstock in the quest to decarbonize the fuels and chemical industries. In this way, it is imperative that researchers prioritize efforts for increased chances of economic viability of the microalgae biomass supply chain. This work applies a sequential techno-economic assessment (TEA) methodology to compare the economic trade-offs between microalgae growth and compositional shift while solving for an alternative economic metric: intrinsic algal biomass value, i.e., the revenue obtained from processing and converting the main components of algae biomass - carbohydrates, lipids, and protein - to a portfolio of products. … | (microalgae biomass, economic trade-off, non-medical) |
Rational engineering of gas-fermenting bacteria for high yields of bioproducts is vital for a sustainable bioeconomy. It will allow the microbial chassis to renewably valorize natural resources from carbon oxides, hydrogen, and/or lignocellulosic feedstocks more efficiently. To date, rational design of gas-fermenting bacteria such as changing the expression levels of individual enzymes to obtain the desired pathway flux is challenging, because pathway design must follow a verifiable metabolic blueprint indicating where interventions should be executed. Based on recent advances in constraint-based thermodynamic and kinetic models, we identify key enzymes in the gas-fermenting acetogen Clostridium ljungdahlii that correlate with the production of isopropanol. … | (gas-fermenting, bioproduction, non-medical) |
A long-standing goal of machine-learning-based protein engineering is to accelerate the discovery of novel mutations that improve the function of a known protein. We introduce a sampling framework for evolving proteins\n <italic>in silico</italic>\n that supports mixing and matching a variety of unsupervised models, such as protein language models, and supervised models that predict protein function from sequence. By composing these models, we aim to improve our ability to evaluate unseen mutations and constrain search to regions of sequence space likely to contain functional proteins. Our framework achieves this without any model fine-tuning or re-training by constructing a product of experts distribution directly in discrete protein space. … | (protein engineering, machine learning, non-medical) |
Interdependence across time and length scales is common in biology, where atomic interactions can impact larger scale phenomenon. Such dependence is especially true for a wellknown cancer signaling pathway, where the membrane-bound RAS protein binds an effector protein called RAF. To capture the driving forces that bring RAS and RAF (represented as two domains, RBD and CRD) together on the plasma membrane, simulations with the ability to calculate atomic detail while having long time and large length- scales are needed. The Multiscale Machine-Learned Modeling Infrastructure (MuMMI) is able to resolve RAS/RAF protein–membrane interactions that identify specific lipid–protein fingerprints that enhance protein orientations viable for effector binding. …. | (Protein binding, Membrane interactions, Medical) |
Here we have reported novel polymeric devices to control fluid flow using surface forces and capillary action to facilitate counter-diffusive, on-chip protein crystallization. | (Fluid flow, Protein Crystallization, Non-Medical) |
Each paper has two keywords automatically extracted using OpenAI’s API. For each cluster calculated via a UMAP of the paper’s abstract (using OpenAI’s embeddings), we can then extract the most common keywords. We count the number of occurrences for each keyword per cluster, and also how many papers are in each cluster.
Top 10 Keywords | Occurrences of Top 10 | Number of Papers |
---|---|---|
protein structure, molecular dynamics, protein design, machine learning, bioinformatics, protein binding, deep learning, protein structure prediction, protein dynamics, protein folding | 104 | 239 |
photosynthesis, light harvesting, photocatalysis, solar fuel, biohybrid, plasmonics, electron transfer, thylakoid membrane, energy transfer, optogenetics | 55 | 56 |
biofuel production, biomass conversion, cellulases, lignocellulose degradation, biomass deconstruction, biofuel, biotechnology, pretreatment, enzyme engineering, bioenergy | 48 | 102 |
x-ray crystallography, xfels, crystallography, protein crystallography, xfel, sfx, x-ray diffraction, x-ray scattering, femtosecond crystallography, radiation damage | 44 | 70 |
proteomics, mass spectrometry, hplc, single-cell proteomics, esi-ms, soil microbiome, proteome extraction, forensics, gradient elution, msi | 42 | 47 |
electron transfer, biocatalysis, nitrogenase, enzyme engineering, electrocatalysis, bioelectronics, enzymatic catalysis, artificial metalloenzymes, synthetic catalysis, electrochemistry | 41 | 73 |
metabolic engineering, synthetic biology, biofuel production, fermentation, microbial biosynthesis, cell-free systems, biotechnology, metabolic pathways, methane conversion, methylotrophy | 40 | 68 |
nanomaterials, self-assembly, nanoparticles, protein assembly, biomineralization, biomimetics, protein self-assembly, bioactivity, protein nanofiber, nanoparticle assembly | 34 | 95 |
crispr, crispr-cas, genome editing, crispr/cas9, cas9, genetic engineering, epigenetics, genome engineering, anti-crisprs, gene expression | 33 | 32 |
microalgae, biofuel production, algal biofuel, biorefinery, algal cultivation, hydrothermal liquefaction, biofuel, bioenergy, htl, biomass conversion | 25 | 38 |
bmcs, metabolic engineering, synthetic biology, carboxysomes, microcompartment assembly, self-assembly, microcompartments, carboxysome, bionanoreactors, electron transfer | 25 | 25 |
synthetic biology, transcription factors, gene expression, transcription factor, transcriptional regulation, rna transcripts, translation control, standardization, cell cycle, ode | 22 | 25 |
fluorescence, protein modification, photoswitching, fret, imaging, protein engineering, gfp, fluorescence microscopy, fluorophore, click chemistry | 22 | 29 |
lipid bilayer, protein phase separation, biomolecular condensation, lipid rafts, membrane proteins, transmembrane proteins, biomimetics, protein-lipid interactions, serum solutions, lipid self-assembly | 18 | 26 |
photosynthesis, water-use efficiency, nswer: (acetate metabolism, light environment, cam photosynthesis, cam, carbon cycling, plant traits, crop physiology, carbon concentrating mechanism | 17 | 12 |
ion transport, microtubules, ion permeation, microfluidics, ion affinity, nsss, polymer partitioning, nanosize channels, tau, myofilament sensitivity | 16 | 29 |
enzyme engineering, protein dynamics, zinc transport, heme peptides, toxicology, organophosphorus, opaa, acyl-enzyme complex, nswer: (beta-lactamase, halophiles | 14 | 16 |
self-assembly, nanoscale engineering, protein cages, vlp, plasmonics, virus nanotechnology, viral capsids, protein nanostructures, vlp self-assembly, enzyme encapsulation | 14 | 13 |
signal transduction, protein binding, optogenetics, gpcr, nmda receptors, phototransduction, cell replacement therapy, eywords: (neuron transplantation, fcεri signaling, protein interactions | 14 | 17 |
metabolic engineering, nswer: (metabolic engineering, systems biology, targeted proteomics, bioengineering, systems modeling, proteomics, protein post-translational modifications, macromolecular expression, metabolism | 14 | 13 |
biosynthesis, polyketides, nswer: (polyketide synthesis, nswer: (fas ii, menaquinone biosynthesis, antibiotic-development, fungal biosynthesis, natural products, o-methylation, clickable functionality | 14 | 19 |
biomolecules, fluorescence microscopy, electron microscopy, image processing, nano-particles, adaptive optics, 3d tracking, spt, photosynthesis, electron transfer | 14 | 22 |
metabolic engineering, biomass utilization, nswer: (anaerobic catabolism, phenylpropanoid biosynthesis, aromatic acid decarboxylation, biosynthesis, 2-pyrrolidone, muconic acid, biomass conversion, tca cycle | 13 | 14 |
protein binding, drug design, protein methylation, nswer: (braf kinase, nswer: (protease engineering, peptide assembly, catalytic efficiency, target selectivity, bromodomain inhibition, melanoma | 13 | 42 |
ncaas, protein engineering, genetic engineering, insecticide, herbicides, codon usage, recombinant gene technologies, cry toxins, pore-forming toxins, lygus species | 13 | 20 |
antibody engineering, fluorescence chimeras, protein binding, scfv, protein stability, antibody recognition, nswer: (epigenetics, antibody selection, hapten detection, fc flexibility | 12 | 12 |
vaccines, foot-and-mouth disease, vp24 protein, flavivirus, vaccine, nswer: (fmdv, drug development, hcv, hiv, eywords: (ribosomal frameshifting | 12 | 13 |
plant transformation, plant breeding, forestry, rice genetics, salt tolerance, agrobacterium, gene flow, switchgrass, biomass, alfalfa | 11 | 13 |
protein assembly, energy storage, supramolecular polymerization, fmoc-ff, molecular gels, nanostructure control, polymer chemistry, nanoparticles, liquid repellency, peptoids | 11 | 39 |
protein engineering, bioconjugation, protein binding, riboswitch, cyclic dinucleotide, double membrane, type iii secretion system, sortase enzymes, pilus biogenesis, dna binding | 11 | 11 |
nswer: (biomass conversion, bioproducts, algae biotechnology, sustainable products, nswer: (camelina, seed metabolism, protein secretion, chlamydomonas reinhardtii, oil biosynthesis, seed development | 10 | 21 |
cell-free protein synthesis, spatial organization, protein expression, rna binding, nswer: (translation regulation, transcriptional regulation, hiv ltr, resource allocation, protein regulation, plant biology | 10 | 11 |
The NSF and the DoE data analyzed here are not the only sources of information to mine.