Dalbavancin Limitations in accuracy and True Positive Rates
Limitations in accuracy and True Positive Rates of biomarker predictions with existing methods and models, may have different causes. Cellular lipid profiles are very complex , and currently incompletely represented in the human genome-scale models. Since many flavoprotein-related diseases affect lipid metabolism it is likely that a better representation of the lipid metabolism will improve the predictions. Additionally, our new method of cofactor implementation could be extended to account for all different cofactors required in human metabolism. Future improvement of our method may involve a differentiation in stoichiometric coefficients of flavin usage per enzyme depending on the specific protein half-life. This would allow the incorporation of differences in efficiency of flavin utilization for various flavin-dependent enzymes, thereby increasing the accuracy of metabolic predictions. In addition, one should remain critical on our assumption that all FAD or FMN is tightly bound as a prosthetic group. While some flavoproteins have FAD covalently bound, most have a non-covalent, yet tight-binding FAD or FMN. Some flavoproteins, however, may have a relatively low FAD binding affinity. This holds for instance for bacterial two-component monooxygenases in which reduced FAD must be translocated from one protein domain to another . Low FAD affinity of cancer-associated variants of NAD(P)H quinone oxidoreductase 1 leads to low protein stability . We are not aware of low-affinity flavoproteins that depend on free Dalbavancin of reduced FAD.
We noted that the currently most extensive reconstruction of human metabolism, Recon 3D, showed a significant decline in the number of correctly predicted biomarkers compared to its predecessors (Table S5). By extending the coverage of the metabolic network, alternative pathways have been created. One may hypothesise that their physiological relevance is smaller in reality than in the model, e.g. due to kinetics, spatial separation, or thermodynamics. This limitation can possibly be overcome by using tissue-specific models with an appropriate set of boundaries for the exchange reactions, as has been proposed recently by Thiele et al. . Finally, it is quite likely that some biomarkers will only be predicted correctly when kinetic and thermodynamic constraints are included.
The following are the supplementary data related to this article.
Sources of funding This work was supported by the Marie Curie Initial Training Networks (ITN) action PerFuMe [project number 316723] and the University Medical Center Groningen. BMB was further supported by a CSBR grant from the Netherland Organization for Scientific Research (NWO) supporting the Systems Biology Centre for Energy Metabolism and Ageing [853.00.110].
CRediT authorship contribution statement Agnieszka B. Wegrzyn: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Validation, Visualization, Writing – original draft; Sarah Stolle: Conceptualization, Validation, Writing - review & editing; Rienk A. Rienksma: Methodology, Writing - review & editing; Vitor A.P. Martins dos Santos: Funding acquisition, Project administration, Writing - review & editing; Barbara M. Bakker: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Writing - review & editing; Maria Suarez Diez: Conceptualization, Methodology, Software, Supervision, Writing - review & editing.
Introduction Concern about the relationship between diabetes and cognitive impairment has grown in recent years, and dementia has been referred to as a diabetes complication [, , , ]. Hyperglycaemia is a primary pathological factor that is responsible for numerous diabetic complications. Hyperglycaemia may be the primary initiator of diabetic complications and glycaemic control is achieved using medication; however, the progression of diabetic complications persists after complete glucose normalization, which supports a phenomenon called metabolic memory . Metabolic memory refers to diabetic stresses that persist after glucose normalization, and hyperglycaemia and metabolic memory are major factors for diabetes complications . An initial report almost 30 years ago suggested the existence of metabolic memory, which was responsible for the progression of incipient diabetic retinopathy during good glycaemic control in diabetic dog models . Metabolic memory was widely studied in different types of models related to diabetic complications, such as diabetic nephropathy  and diabetic retinopathy  in vivo and vascular smooth muscle cells  and endothelial cells  in vitro. Dementia and neurodegenerative disorders are complications of diabetes, which may be mediated via metabolic memory . However, our understanding of the mechanisms underlying the relationship of metabolic memory to central nervous diseases is not complete, and effective therapeutic strategies are urgently needed. Our understanding of the molecular mechanisms underlying metabolic memory is limited, but evidence supports that advanced glycation end products (AGEs) and reactive oxygen species play essential roles . The binding of AGEs to the receptor for AGEs elicits abnormal reactive oxygen species generation, and reactive oxygen species accumulation also excessively activates the AGEs/the receptor for AGEs axis . This activation leads to a self-maintaining vicious loop that is independent of glucose level and results in the target organ damage that is responsible for diabetic complications . Therefore, the addition of potential agents may be beneficial in the treatment of AGEs- and reactive oxygen species-mediated damage and abnormalities, in addition to glucose normalization, to ameliorate diabetic complications, including dementia.