NEGATIVE DATA™
Negative Data™ is a journal for the days of generative artificial intelligence (AI)-assisted publication
One of the major problems with AI-assisted scientific publications is that AI generally cannot distinguish between correct (reproducible) and incorrect (unreproducible) data obtained in the past and at present. Negative Data is a journal that explores whether AI can reduce the risks of incorporating incorrect information in the corresponding publications.
Journal 1
JOURNAL OF NEGATIVE DATA IN IMMUNOHISTOCHEMISTRY
The weakness of AI lies in judging whether a given biomedical dataset is reproducible or not. When necessary, we as experimental scientists try to replicate such a data set using as many positive and negative controls as possible. Experiments without relevant control(s) are often misleading. Typical examples are the results obtained by immunochemical methods. For instance, the conventional immunohistochemical observations of postsynaptic glutamate receptor and neuroligin 1 turned out to be non-specific signals when corresponding knock-out mice were used as negative controls(Watanabe et al., 1998; Zhang et al., 2015a).
Journal 2
JOURNAL OF NEGATIVE DATA IN ALS RESEARCH
Pharmacological candidates developed to treat ALS all failed when mice overexpressing mutant superoxide dismutase 1 (SOD1) were used as “gold-standard” models (Renton et al., 2014). It was discovered that an abnormality in SOD-1 is not a major pathology of familial and sporadic ALS cases (Ajroud-Driss and Siddique, 2015) (Saberi et al., 2015) (Frakes et al., 2014).