OpenLab - Compound_CalibAmount in table but displayed for sample not for calibration standard

You can set up this calculation with IR or a CC file, but you have to make sure you are using the right fields and your bracketing mode is right for the calibration. If you test the fields in IR you can find the one that automatically calculates concentrations of compounds off calibration curves. Just note that it will only work if the samples are after the curve or you set the bracketing mode to overall (or similar). Then just divide by the input concentration and multiply by 100.

Thanks, but you didn't understand.

Bracketing has nothing to do it with that - I want to run sample and check results against std concentrations. All I need is to display STD concentrations in table for each compound.

Tom

Which STD concentration? You have six different levels here. If you want to show all of them in a separate table, you can use a default calibration table and plot in a report. If you want them in the same table, then you need to make a separate custom field that connects the sample to the right calibration level unless the prepared sample concentration is input as exactly the same as the corresponding level concentration.

Can you do a sketch of what you want exactly? It would help me understand what you are trying to achieve.

I want to have report for SAMPLE in which I will be able to display STD concentrations for relevant compounds - those that you put in processing method.

But I want to do it only based on processing method that it is used to process sample - in the sequence there will be no STD. I will use saved processing method to analyse SAMPLE.

I hope it is clear.

Thanks

Tom

That's slightly problematic. I know that the system stores the calibration of compounds because if you don't clear calibration the new one will include old data, but I'm not sure if it will properly apply to samples if there are no calibration standards within the sequence.You can attempt to find out if it is calculating concentrations still by referencing the compound concentration fields. If it is, that would allow you to calculate the accuracy of the calibration being used at least, though without the STD concentrations being stated.

Further, in IR all the lines in the data relate to specific samples, so having no samples that are calibration standards will probably make it impossible to reach STD concentrations unless you find a way to directly access the processing method information from within the reports, which I've never attempted and I don't recall any obvious ways to do that.You can get some very basic information like processing method name and date, but I've not seen a way to get much further than that.

However, you can always create a new data set that includes the calibration data you are attempting to calculate off of. Alternatively, you can just load up both the calibration sequence and the sequence you are looking at and run them both through a cross-sequence report.

I will note that there is a good reason that calibrations are generally done within the same sequence though. You may never know when the instrument craps out on you and in an assay that is done over and over again you can encounter drift. For this reason, it's safer documentation-wise to always include your calibration in the same sequence. It makes justifying your measurements much easier for regulatory bodies. How are they to know that you don't have 10 different calibration sequences you are picking from to get the results you want?

ColeFunk said:

I will note that there is a good reason that calibrations are generally done within the same sequence though. You may never know when the instrument craps out on you and in an assay that is done over and over again you can encounter drift. For this reason, it's safer documentation-wise to always include your calibration in the same sequence. It makes justifying your measurements much easier for regulatory bodies. How are they to know that you don't have 10 different calibration sequences you are picking from to get the results you want?

Sometimes STD preparation is quite time consuming. Running control sample against saved calibration curve is quick check how "good" calibration is and if it still shows acceptable result for control sample (for that I will use saved processing method). Therefore I will run sequence with sample only (which technically is the same concentration level as one of the std used for calibration curve preparation) and I will check sample result against specific std level (which has the same concentration). So I will need somehow to have in table results real STD concentration, so I can compare results. The only way to do it is to have concentrations from processing method.

Believe me - I', doing it routinely in excel but I was hoping to do it in OL

Thanks

TOm

ColeFunk said:

I will note that there is a good reason that calibrations are generally done within the same sequence though. You may never know when the instrument craps out on you and in an assay that is done over and over again you can encounter drift. For this reason, it's safer documentation-wise to always include your calibration in the same sequence. It makes justifying your measurements much easier for regulatory bodies. How are they to know that you don't have 10 different calibration sequences you are picking from to get the results you want?

Sometimes STD preparation is quite time consuming. Running control sample against saved calibration curve is quick check how "good" calibration is and if it still shows acceptable result for control sample (for that I will use saved processing method). Therefore I will run sequence with sample only (which technically is the same concentration level as one of the std used for calibration curve preparation) and I will check sample result against specific std level (which has the same concentration). So I will need somehow to have in table results real STD concentration, so I can compare results. The only way to do it is to have concentrations from processing method.

Believe me - I', doing it routinely in excel but I was hoping to do it in OL

Thanks

TOm

Hello,

You cannot use the calibration table in calculations unless you have a cal std in the result set. You will need to use a compound custom field to enter the calibration values for the check injections in this case. You could use a CC constants file as well if the standard values do not change often. 

Marty Adams

Thanks. I will try CC. Will it work with built-in report as custom field or I need to create new report template from scratch?

Tom

On your other question from the most recent comment, you can access CC file variables from the report, but they likely won't do it for you. You won't have to start from scratch on a brand new report, but just modify an existing report to include those constants where you want and save it as a different report to preserve the default. If you want to show them inline to a table, you will likely have to create a custom calculation that links the concentration you want to a new variable keyed by sample_ID and have some criteria to limit this to only the control sample. Then you call that variable in a specific column of the table. If you aren't familiar with the reports, this may be a bit difficult, but once you know the basics it won't be too hard.

I'm unsure if you will be able to quantify concentrations of samples automatically without any calibration standards in the sequence, but you will certainly be able to access the concentrations of the standards with the CC file. You can try pulling both the calibration data and the sample data using one report if the CC file doesn't give you what you want. It is pretty quick and easy to do. If it wont quantify sample concentrations, you can either merge them into one data set, which will allow it to quantify for sure or you can set up the calculations manually in the report.

I understand why you wouldn't want to make calibration standards every time, but determining if a whole curve is still valid by testing one point may not always give the right answer. For methods that have gone through a whole validation proving linearity with an insignificant y-intercept, it is generally acceptable to switch to a single point calibration, which is about the same as you are talking about, but with a few injections to gather statistical variability. However, the difference here is that the calibration always depends on the calibration in the most recent control samples. If there is drift, the new calibration data aligns with it. Of course, you can calculate however you want. I'm just trying to protect your results from regulatory ridicule. There are multiple issues that can occur over the lifespan of an instrument that can change what the correct calibration curve is and checking one point of the curve may not always tell you when they are happening. Always cover your rear or it may just come back to bite you.

Hello,

I think there are other posts where I go over this in more detail. If you want to do a recovery or check calculation for more than one compound and it is not a reinjection of a standard level or the standards are not in the results set, the easiest way is a compound custom field. You setup a compound custom field in the CDS project and then fill in the values for each compound when running the sequence or in DA. You will need to enter the values for all injections where you are doing the calculations. You can then add a column to your tables where you calculate the error or recovery values you want. 

Marty Adams

=Val(Compound_Amount / CFE((Compound_CustomFields),"QC Conc") * 100)