TY - GEN
T1 - A Unified Benchmark and Conditional Sequence Modeling Framework for Future Influenza Evolution Prediction
AU - Li, Zhuolun
AU - Huang, Xulinyi
AU - Yang, Jiayu
AU - Lin, Qianyao
AU - Wan, Xiaohua
AU - Cui, Yuxiao
AU - Madhi, Shabir
AU - Zhang, Fa
AU - Zhang, Dongxu
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2027.
PY - 2027
Y1 - 2027
N2 - Vaccination remains the most effective strategy for seasonal influenza prevention, but vaccine strain selection depends on anticipating future circulating variants before they are fully observed. This is challenging because influenza viruses evolve rapidly and existing studies often use heterogeneous datasets, targets, and evaluation settings. We formulate future influenza evolution prediction as a conditional sequence modeling problem and present a unified benchmark, with an initial release on influenza A/H1N1 hemagglutinin (HA). The benchmark distinguishes raw and aligned sequence layers, separates hard and soft target definitions, and adopts leakage-safe rolling evaluation. We instantiate the benchmark with a representative conditional seq2seq framework that combines pretrained protein representations, prevalence signals, and a compact multi-objective training objective. The resulting study provides a reproducible basis for systematic comparison in future influenza forecasting research.
AB - Vaccination remains the most effective strategy for seasonal influenza prevention, but vaccine strain selection depends on anticipating future circulating variants before they are fully observed. This is challenging because influenza viruses evolve rapidly and existing studies often use heterogeneous datasets, targets, and evaluation settings. We formulate future influenza evolution prediction as a conditional sequence modeling problem and present a unified benchmark, with an initial release on influenza A/H1N1 hemagglutinin (HA). The benchmark distinguishes raw and aligned sequence layers, separates hard and soft target definitions, and adopts leakage-safe rolling evaluation. We instantiate the benchmark with a representative conditional seq2seq framework that combines pretrained protein representations, prevalence signals, and a compact multi-objective training objective. The resulting study provides a reproducible basis for systematic comparison in future influenza forecasting research.
KW - Benchmark
KW - Influenza forecasting
KW - Multi-objective learning
KW - Protein language models
KW - Time extrapolation
KW - Vaccine strain selection
UR - https://www.scopus.com/pages/publications/105046230331
U2 - 10.1007/978-981-92-3716-6_17
DO - 10.1007/978-981-92-3716-6_17
M3 - Conference contribution
AN - SCOPUS:105046230331
SN - 9789819237159
T3 - Lecture Notes in Computer Science
SP - 211
EP - 221
BT - Bioinformatics Research and Applications - 22nd International Symposium, ISBRA 2026, Proceedings
A2 - Cui, Xuefeng
A2 - Lei, Xiujuan
A2 - Porozov, Yuri
PB - Springer Science and Business Media Deutschland GmbH
T2 - 22nd International Symposium on Bioinformatics Research and Applications, ISBRA 2026
Y2 - 22 July 2026 through 24 July 2026
ER -