KREATiS opened its doors in 2014 as a privately funded Research & Development organisation based close to Lyon, providing endpoints for OECD and EC regulatory studies in a similar way to a Contract Research Organisation. These can be used for regulatory purposes by companies working in the field of chemicals (eg. REACH dossiers). The difference is that the endpoint we provide are predictions derived from in silico models Based on high quality experimental data which means we can deliver almost instantaneous results for just a fraction of the cost of our lab coated colleagues. The team consists of highly experienced QSAR modellers, ecotoxicologists, toxicologists, chemists and a statistics expert. All the High-Accuracy QSARs (HA-QSARs) developed at KREATiS are commercialised as iSafeRat® services (in Silico Algorithms For Environmental Risk And Toxicity). Until 2017 all the iSafeRat® HA-QSARs were only accessible to our clients as part of KREATiS’ expert service, however as a part of initiative to upgrade the way our clients access iSafeRat services, in January 2018 we started to make the iSafeRat® HA-QSARs for physicochemical and ecotoxicological endpoints accessible to registered users via iSafeRat® Online.
Often, the ultimate objective of the QSAR model is to prioritise chemicals for further testing rather than to predict an exact experimental value. In this case, we observe:
The number of results used to develop models is considered more important than the quality of the data used, the relationship(s) between several parameters used is sometimes over-complex and rarely explained, results are provided without reliability and confidence indicators and are therefore suitable for screening but not for regulation purposes.
For the above reasons it seems necessary to distinguish between QSARs created to provide approximations of experimental values and those specifically designed to substitute for specific experimental endpoints. We call these latter models High-Accuracy QSARs or “HA-QSARs”.
Aim of a HA-QSAR is to provide an endpoint value at least as accurate (or even more so, depending on the study in question) as a normalised (guideline) experimental study for a property or activity. Compared to standard QSARs, HA-QSARs offer values reliable enough to replace experimental studies and meet criteria such that they can be used in many regulatory submissions (e.g. for REACH).
KREATiS aims to supersede studies principally by use of high accuracy, in Silico algorithms at a fraction of the price of the laboratory equivalent. This is to ensure that the cost to the client is always lower than the experimental equivalent and in proportion with the research effort required to achieve it. Moreover, we have faith in our models and we want you to as well. For this reason, KREATiS offers a money back guarantee on all of its products. In order to benefit from this offer the following must be met:
*The study must be conducted following best practice and be monitored by KREATiS or a KREATiS referenced scientist or partner
For each substance predicted with iSafeRat® online, the user can download a prediction summary report which contains the predicted value, confidence intervals around it as well as the applicability domain status for each endpoint of user’s interest. In the near future, iSafeRat® online will also allow the users to download a full QSAR Model Report (QMRF) and QSAR Prediction Report (QPRF) necessary for successful REACH submission of your endpoint to ECHA. In the meantime, the users can still order these regulatory reports using the current interface and once the documents are ready, they will be emailed to the user. If required a complete Robust summary for the endpoint can be prepared by KREATiS and will be sent in i5z format.