You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
LC area percent is a common measurement type in industry HTE. In the recent Pfizer dataset ord-data/#189 I used a custom measurement type for LC area percent. There is a purity type which is probably the most accurate description of what is measured, but LC area percent in the context of HTE is generally interpreted as a surrogate for yield.
Describe the solution you'd like
An area_percent measurement type could avoid suggesting the measurement is a yield or purity, and would be complementary to the area/count/intensity types.
This can be discussed at an ORD meeting.
The text was updated successfully, but these errors were encountered:
Is your feature request related to a problem? Please describe.
LC area percent is a common measurement type in industry HTE. In the recent Pfizer dataset ord-data/#189 I used a custom measurement type for LC area percent. There is a purity type which is probably the most accurate description of what is measured, but LC area percent in the context of HTE is generally interpreted as a surrogate for yield.
Describe the solution you'd like
An area_percent measurement type could avoid suggesting the measurement is a yield or purity, and would be complementary to the area/count/intensity types.
This can be discussed at an ORD meeting.
The text was updated successfully, but these errors were encountered: