GNorm2 is a gene name recognition and normalization tool with optimized functions and customizable configuration to the user preferences. The GNorm2 integrates multiple deep learning-based methods and achieves state-of-the-art performance. GNorm2 is freely available to download for stand-alone usage.
Download GNorm2 here
The codes have been tested using Python3.8/3.9 on CentOS and uses the following main dependencies on a CPU and GPU:
Please firstly download
GNorm2 to your local.
Below are the well-trained models (i.e., PubmedBERT/Bioformer) for Gene NER and Species Assignment.
Models for Gene NER:
BioC-XML or
PubTator formats are both avaliabel to GNorm2.
This research was supported by the Intramural Research Program of the National Library of Medicine (NLM), National Institutes of Health.
