Spiked Sample Calibration and quantitation

Hello,

Disclaimer: I am new to the software. I have created a processing method with over 100 potential compounds I want to look for. My samples (known mass) are spiked with a known mass of standard solution containing 3 standards in different concentrations (ppm). The point I am stuck at the moment is is that after setting my samples to spiked, I cannot get the identified peaks to be quantified, and I also have some minor questions regarding the processing method as well. Details are as follows:

Software: OpenLAbs CDS v 2.7

Equipment is a GC-FID/MS to simultaneously quantify and qualify a sample.

What the issue is: 

  • I only see an amount and concentration in the results tab for the 3 standards I've inputted.
  • The calibration curve table or internal standard peaks says "Peak is ISTD", no calibration data abailable
  • For other compounds, the calibration table shows "not quantified", with also no calibration data available.

What I've done so far: 

  • In the injection list, I've set all samples to "spiked". I've added a sample amount (g) to each. ISTD amount 1-3 are blank atm.
  • In the compounds indetifiaction tab, I've assigned my 3 standrads as standards in the compound list and assiciated all the remaining compounds to either of them as associated time reference compunds.. RT update after each run for every comound.
  • In the compound table of the Compounds calibration tab, I have is ISTD selected for my 3 standards, with their concentrations in PPM inputted to ISTD amount. I have assigned all the other compounds to either standards. Mode: curve, Weighing method: none, curve model: linear, Origin: force, response scaling: none.
  • In the general section of the compounds calibration tab, I have internal standard selected. Number of levels: 1. Curve calculation: from average per level. RF definition: Amount per response.  

Additionally, a question about the processing method: out of my 100 potential compounds, some do not occur in all the samples. i've set every compound to adjust for retention time shift based on the shift of my internal compounds in the method. Is there a way to also shift the expected retention times of undetected compounds in my method?

Also, is there a way to automatically identify unidentified peaks in the FID spectra using the MS spectra? As in if a peak is not identified, the program would use the MS data to provide the most likely match?

Thank you in advance for the help.

Parents
  • Hello,

    You only have 3 compounds in your standard, and they are all marked as ISTD? In order to use internal standards, you must have curves for at least one compound that is not marker an internal standard. Setting sample type of spiked does not do anything except add that label and change the sample type for reporting, the processing and calculations are the same as sample.  I am not sure what you mean "undetected compounds in my method" are you talking about missing peaks or unknowns. If you are talking about the RT update for missing peaks, then no, that mechanism only works when peaks are found.  No, you cannot use the FID peaks for MS search. 

    Marty

  • Thank you for the reply, I will have a go at it tomorrow. So essentially I just need to change one of my internal standard compounds to a normal compound, and use that as a single point calibration? It just seemed logical that every internal standard compound serves as its own single-point calibration within a sample, and  I could not find the requirements to make it work. The sample descrition for spiked just seemed to fit what I was looking for.

    It is unfortunate that missing peaks cannot be shifted. I think it would be a usefull feature to add. It will make complete qualitative any quantitative analysis of my samples a bit more time consuming.

    Regarding the MS search, is there any way to merge the MS data with the FID?  Or I can only use the MS to identify a compound, then add the compound to the method, and link it to the FID signal?

    Thank you again for the answer.

  • I think in the next update they've listed that compounds will be cross-signal, but for now each compound can only be in one signal. My lab simply names them as [compound]_[signal] for obvious reasons. Anyway, it's dangerous business assuming that peaks at the same time in different signals are the same compound because there is always a lag between detectors since the compounds must traverse the physical space between said detectors. How big the lag is will be dependent on your flow rate, but you can estimate based on your ISTDs. Also, there's always a chance that compounds that are readily detected in one signal are not so in the other and a worse scenario is when you have two compounds that elute at similar times, but one is detected in one signal while the other is in the other. Especially with very nonspecific detectors like FID, it would be hard to confirm that this isn't the case. The only way you could really confirm is to run a standard of the expected compound and do some sort of calibration in each detector to make sure the predicted concentrations are consistent between signals. I highly doubt you would want to do this for up to 100 compounds, I sure wouldn't.

    Anyway, to answer your question about the compounds being assigned in FID based on MS, I highly doubt that is a feature because, again, in the current version you have (same as mine) compounds (as stored in the data) can only be in one signal.

    Also, for quantifying compounds other than your internal standards entering information into the calibration table tells the software that you are running a calibration for that compound. Meaning, you have at least one level that you know the concentration of in your "Cal. Std." samples. This isn't the case for you as far as I could tell. I don't believe there is a feature that easily allows you to quant unknowns off of knowns, but you can absolutely set this up in the Intelligent Reporter. Below is how I would do this:

    1. Create three variables in IR (one for each ISTD) that will be used to store the RF of the associated ISTD, called IS1RF, IS2RF, and IS3RF for names of IS1, IS2, and IS3 in this list of steps. Make the Unique Key Value = Signal_Name and formula = Avg(Compound_AreaSum/Compound_CalibAmount). Group the objects for the variables separately. For each, set the group repeat =1. Then filter by Sample_Type=1 and Compound_Name =[the compound name of one of the ISTDs]. deselect the checkbox to make it visible.

    2. Add a compound custom parameter to the project called "IS Name". Fill in the IS Name for each unknown in this parameter in the injection list (do it in one, then copy and paste the table for the rest). Leave this space empty for the ISTDs or fill in with their own name.

    3. Make a table (I recommend going with the Compound Summary default and just adding/removing columns as necessary). In the filtering tab of this table, filter by Expression = len(CFE((Compound_CustomFields),"IS Name")), Operation = >, Value = 0.

    4. Change the "Concentration" column's expression to =Compound_AreaSum/IF(CFE((Compound_CustomFields),"IS Name")="IS1", IS1RF(Signal_Name), IF(CFE((Compound_CustomFields),"IS Name")="IS2", IS2RF(Signal_Name), IF(CFE((Compound_CustomFields),"IS Name")="IS3", IS3RF(Signal_Name),1)))

    This should give you a table for each compound listing all injections with the concentration quanted off their respective ISTDs. If you entered something incorrectly, it gives just the plain area, which should be easy to identify if you leave area as a column in the table (note: all text here is case-sensitive). There are more elegant and flexible ways to do this, but this is the path I saw to solve your problem with the fewest steps and least background knowledge.

    Hope this helps,

    Cole

Reply
  • I think in the next update they've listed that compounds will be cross-signal, but for now each compound can only be in one signal. My lab simply names them as [compound]_[signal] for obvious reasons. Anyway, it's dangerous business assuming that peaks at the same time in different signals are the same compound because there is always a lag between detectors since the compounds must traverse the physical space between said detectors. How big the lag is will be dependent on your flow rate, but you can estimate based on your ISTDs. Also, there's always a chance that compounds that are readily detected in one signal are not so in the other and a worse scenario is when you have two compounds that elute at similar times, but one is detected in one signal while the other is in the other. Especially with very nonspecific detectors like FID, it would be hard to confirm that this isn't the case. The only way you could really confirm is to run a standard of the expected compound and do some sort of calibration in each detector to make sure the predicted concentrations are consistent between signals. I highly doubt you would want to do this for up to 100 compounds, I sure wouldn't.

    Anyway, to answer your question about the compounds being assigned in FID based on MS, I highly doubt that is a feature because, again, in the current version you have (same as mine) compounds (as stored in the data) can only be in one signal.

    Also, for quantifying compounds other than your internal standards entering information into the calibration table tells the software that you are running a calibration for that compound. Meaning, you have at least one level that you know the concentration of in your "Cal. Std." samples. This isn't the case for you as far as I could tell. I don't believe there is a feature that easily allows you to quant unknowns off of knowns, but you can absolutely set this up in the Intelligent Reporter. Below is how I would do this:

    1. Create three variables in IR (one for each ISTD) that will be used to store the RF of the associated ISTD, called IS1RF, IS2RF, and IS3RF for names of IS1, IS2, and IS3 in this list of steps. Make the Unique Key Value = Signal_Name and formula = Avg(Compound_AreaSum/Compound_CalibAmount). Group the objects for the variables separately. For each, set the group repeat =1. Then filter by Sample_Type=1 and Compound_Name =[the compound name of one of the ISTDs]. deselect the checkbox to make it visible.

    2. Add a compound custom parameter to the project called "IS Name". Fill in the IS Name for each unknown in this parameter in the injection list (do it in one, then copy and paste the table for the rest). Leave this space empty for the ISTDs or fill in with their own name.

    3. Make a table (I recommend going with the Compound Summary default and just adding/removing columns as necessary). In the filtering tab of this table, filter by Expression = len(CFE((Compound_CustomFields),"IS Name")), Operation = >, Value = 0.

    4. Change the "Concentration" column's expression to =Compound_AreaSum/IF(CFE((Compound_CustomFields),"IS Name")="IS1", IS1RF(Signal_Name), IF(CFE((Compound_CustomFields),"IS Name")="IS2", IS2RF(Signal_Name), IF(CFE((Compound_CustomFields),"IS Name")="IS3", IS3RF(Signal_Name),1)))

    This should give you a table for each compound listing all injections with the concentration quanted off their respective ISTDs. If you entered something incorrectly, it gives just the plain area, which should be easy to identify if you leave area as a column in the table (note: all text here is case-sensitive). There are more elegant and flexible ways to do this, but this is the path I saw to solve your problem with the fewest steps and least background knowledge.

    Hope this helps,

    Cole

Children
No Data
Was this helpful?