| Design | Est. half-life (hrs)a | AUPinit. 14 ntsb | AUPCDSc | dG(MFE) (kcal/mol)d | CAI |
|---|---|---|---|---|---|
| Putative BioNTech/Pfizer_BNT-162b2e | 0.21 | 0.50 | 0.40 | -1341.1 | 0.95 |
| Putative Moderna mRNA-1273e | 0.22 | 0.43 | 0.39 | -1510.1 | 0.98 |
| IDT codon optimization | 0.22 | 0.46 | 0.51 | -1089.5 | 0.73 |
| GENEWIZ codon optimization | 0.25 | 0.58 | 0.49 | -1304.3 | 0.95 |
| GC-richf | 0.27 | 0.63 | 0.42 | -1617.2 | 0.80 |
| LinearDesigng | 0.32 | 0.42 | 0.20 | -2533.3 | 0.72 |
| Superfolder-v1 | 0.34 | 0.86 | 0.22 | -2382.1 | 0.73 |
| Superfolder-v2 | 0.34 | 0.90 | 0.22 | -2375.4 | 0.72 |
| Superfolder-Delta | 0.33 | 0.86 | 0.23 | -2323.6 | 0.72 |
| Superfolder-Delta-PSU | 0.55h | 0.90 | 0.29 | -1711.2 | 0.64 |
| Superfolder-Omicron | 0.35 | 0.42 | 0.23 | -2278.6 | 0.72 |
| Superfolder-Omicron-PSU | 0.55h | 0.84 | 0.29 | -1712.5 | 0.64 |
aHalf-life estimated from DegScore, machine-learning model for predicting degradation more accurately than average unpaired probability (Leppek, 2021) (see note below). Half-life estimate is for accelerate degradation conditions mimicking cationic environment (10 mM MgCl2, Na-CHES pH 10.0, 24 °C); bAverage unpaired probability of the first 14 nucleotides of the coding sequence (Kozak, 1990) -- should be high to optimize translation; cAverage unpaired probability over entire coding sequence (Wayment-Steele, 2020b) ddG(MFE) calculated in LinearFold-Vienna; eJeong, 2021.; fEach codon is randomly sampled from the most GC-rich codons for that amino acid (Thess, 2015); gZhang, 2020; hUsing PSU heuristic, setting DegScore for U to 0 (see DegScore repository.).
The Superfolder-v2 construct was optimized by Eterna participants to reduce predicted degradation. It is an adaptation of Superfolder-v1, which had been optimized in RiboTree both to minimize DegScore, maximizing in vitro stability, while keeping the first 14 nucleotides of the coding sequence unpaired.
The Superfolder-Delta-PSU and Superfolder-Omicron-PSU constructs were designed from scratch by Eterna participants in the Delta and Omicron challenges, respectively, optimizing a modified DegScore that increases the stability of PSU by setting U degradation to zero.
The Superfolder-Delta construct was developed using the mRNA-hotfix tool: this tool enumerates all GC-rich codon substitutions that result in the Delta mutant spike protein, evaluates their DegScore, and selects the variant with the lowest DegScore.
To calibrate the DegScore model to measured degradation rates on full-length mRNAs, we performed a linear fit between predicted DegScore values for ~233 mRNAs of varying length (Leppek, 2021), as well as the "Roll-Your-Own Structure" dataset. The below plot shows the resulting fit of the degradation rate, and the corresponding half-life.
Specifically, estimated degradation rate (est_k_deg) is calculated as
est_k_deg = m * degscore + b
m = 0.002170959651184987
b = 0.05220886935630193
And estimated half life is calculated as
est_half_life = ln(2) / est_k_deg.
Jupyter notebook to reproduce the above analysis is here.
Eterna participants have designed S-2P and Delta variant mRNAs with a diversity of structures in OpenVaccine project.
In December 2020, 181 S-2P constructs were submitted and voted upon by Eterna participants. The top 9 candidates are included in superfolders_ALL.csv.
Between May and August 2021, 1,563 B.1.617 and Delta variant designs were submitted and voted upon by Eterna participants. The top 8 candidates for both unmodified nucleotides and PSU DegScore are included in superfolders_ALL.csv.
analysis/analysis.py reproduces the above calculations. Usage:
python analysis.py input.csv
Where input.csv is of the form
Designer,sequence
Design_A,AUGCAUAAAUGGUCUUGA
Design_B,AUGCACAAGUGGAGCUGA
Dae-Eun Jeong, D.E. Matthew McCoy, M., Karen Artiles, K., Orkan Ilbay, O., Fire, A., Nadeau, K., Park, H., Betts, B., Scott Boyd, S., Hoh, R., Shoura, M. Assemblies of putative SARS-CoV-2 spike encoding mRNA sequences for vaccines BNT-162b2 and mRNA-1273.
Kozak M. Downstream secondary structure facilitates recognition of initiator codons by eukaryotic ribosomes (1990). Proc Natl Acad Sci U S A. 87(21):8301-5. doi: 10.1073/pnas.87.21.8301
Leppek, K., Byeon, G.W., Kladwang, W., Wayment-Steele, H.K., Kerr, C., ... Barna, M., Das, R. (2021). Combinatorial optimization of mRNA structure, stability, and translation for RNA-based therapeutics. https://www.biorxiv.org/content/10.1101/2021.03.29.437587v1.
Wayment-Steele, H.K., Kim, D.S., Choe, C.A., Nicol, J.J., Wellington-Oguri, R., Sperberg, R.A.P., Huang, P., Eterna Participants, Das, R. (2020). Theoretical basis for stabilizing messenger RNA through secondary structure design. bioRxiv, 262931.
Wayment-Steele, H.K., Kladwang, W., Eterna Participants, Das, R. (2020). RNA secondary structure packages ranked and improved by high-throughput experiments. bioRxiv, 124511.
Thess, A., Grund, S., Mui, B. L., Hope, M. J., Baumhof, P., Fotin-Mleczek, M., & Schlake, T. (2015). Sequence-engineered mRNA without chemical nucleoside modifications enables an effective protein therapy in large animals. Molecular Therapy, 23(9), 1456-1464.
Zhang, H., Zhang, L., Li, Z., Liu, K., Liu, B., Mathews, D. H., & Huang, L. (2020). LinearDesign: Efficient Algorithms for Optimized mRNA Sequence Design. arXiv preprint arXiv:2004.10177.
For answers to any additional questions that might help accelerate the end of the COVID-19 pandemic, please contact Rhiju Das, Stanford University, [email protected].