Deep learning for protein prediction
Title: |
Deep learning for protein prediction |
DNr: |
SNIC 2021/5-297 |
Project Type: |
SNIC Medium Compute |
Principal Investigator: |
Arne Elofsson <arne@bioinfo.se> |
Affiliation: |
Stockholms universitet |
Duration: |
2021-07-01 – 2022-07-01 |
Classification: |
10203 10610 10601 |
Homepage: |
http://bioinfo.se/ |
Keywords: |
|
Abstract
Machine Learning has had profound impact on life-science, as
well as many other scientific areas. Lately, deep learning
strategies have shown great progress in areas such as speech and
image recognition. Here, these methods clearly outperform
earlier machine learning methods. Deep learning methods are also
starting to make an impact in life science, but adaptation has
been slow. We believe a much wider life science community can
start using these tools - and that the impact of this might be
significant. We have a long history in developing widely
adapted bioinformatics methods, primarily within the area of
protein structure prediction. These methods have been based on
deep biological insights and various machine learning metods,
including deep learning. Here, we apply for computational
resource to continue this development.