Notebook link where you can find all the graphs.
Data created on 17.10.2022 at 19:32:45
Data updated on 17.10.2022 at 19:32:45
Support by Ontologies
While searching in NMR-related ontologies, finding terms about the temperature through an NMR assay wasn't possible. When saying "Temperature" here, we mean the temperature at which the NMR spectroscopy was conducted. Unit-wise, one can easily use ontology-driven units.
Data Sanitisation and Missing Values
|Field Type||Field Name||Values Readability||Unit||Missing||Comment|
|MTBLS||dedicated||Temperature||float||separate field||The field is not provided; or the unit was not found in the list of units we provided; or the value is provided as N/A or other similar expressions; or the study "assays" value is "null".||The provided list of units is ['UO:kelvin', 'UO:kelvin:K','Kelvin','kelvin', 'degree Celsius','celsius', 'degree celsius']|
|MW||dedicated||Temperature||machine-readable||integrated||The field is not provided; or an expression of numbers was not found in the field; or the value is provided as N/A or other similar expressions; or the unit (words with c or d for Celsius and k for Kelvin) was not found; or decoding the JSON file that contains the study details has failed due to syntax error there.|
|NMRShiftDB||dedicated||cml:temp||float||separate field||none||The unit is ontology-driven as it is always Kelvin|
|MW||["4C", "4 oC", "281 ± 0.5 K", "15 celsius", "23d", "22 degree Celsius", "27", "5mm", "295.9", "36,85"]||"288", "293"|
|NMRSiftDB||["243","300.0136", "Unreported", "Unknown"]||"288", "293"|
The temperature was easy to obtain in all the three repositories that provide it. However, the inconsistency encountered in the values and units made it time-consuming to harmonize the data. In the end, having units coming (or possibly to be mapped) from ontologies made most of the data clear and machine-readable. The temperature provided in Celsius was converted to Kelvin, and with visualization, one can see that most of the assays were held around room temperature.
Still, some of the values were reported wrongly in Kelvin, as values like 0 and 37 can be seen in the graph showing the number of studies for each temperature (with a logarithmic scale).