Method Validation - Science topic

What was the N for your RNA seq, and qPCR analysis? Surprisingly, the trend is the opposite, which shouldn't be the case, unless the primer isn't specific and amplifies some other segment.

All parameters as per ICH Q2(R1) guideline

Dear Anna Turbanova ,

Interesting question. I agree with Andrei Blasko and Saeed Ayaz Khan . However, you did not mean a regular analytical method validation. You are dealing with dissolution kinetics. The dissolution itself will depend on the physicochemical parameters you used, so the kinetics.

You find in the literature different behavior of the same AFI, for instance losartan, in several mediums, pH, ionic strength, and solvation conditions. Therefore several other surface and chemical interactions may influence the kinetics, not the determination method itself.

For your safety, check the validation with the already used media. The kinetics of significant different media will change. Therefore, the dissolution kinetics may have different parameters that may affect or not the determination method.

Best regards,

WNM

Probably not advisable to continue until you know why you don't get good recovery from the CRM

If you continue as you are, you may be making a serious error in the extraction process, or sampling errors might be affecting the calculations, or the measurement technique might be faulty. You won't know what the "true" results are until you have improved the entire process to within the acceptable tolerance limits.

Your question does leave a few gaps for a complete answer though. A few ideas to help:

If you can go through your method and find a few key points to describe where your method is going awry, then we might be able to provide some more helpful suggestions than these generalised responses.

Using polynomial equations is pretty normal in the pharmaceutical field. Especially when you work with biotech/biologics or detectors with limited range of linearity like ELSD. I have never had compliance issues after submitting analytical methods with polynomial curves.

As hinted above though, you could adjust your sample concentrations to fit within the linear portion of the curve if that's possible for you.

As for the second part of the question, how to justify it. Run a method qualification by doing 5 solutions with concentration ranging from at least 80-120% of your target. Include all your analytes. Run those in triplicates and calculate the %RSD and the regression coefficient. Ideally, R squared has to be 0.99 or more, but for many methods 0.98 is fine. After that, perform accuracy from spiked samples: Spike your diluent/placebo/matrix with your analytes in at least 80-120% range from the target and analyze in triplicates. Calculate recovery (expected ~95-105% for assays) and %RSD. If you don't have matrix other than your diluent, than you can calculate the accuracy from your linearity solutions. Have another colleague run the method on a different day to confirm your results (you can run just linearity curve and one sample at 100% target). This will prove that you have data for your curve and that the method is accurate for quantitative purposes. Then you can propose this to a more regulated environment like GDP or GMP lab to do full method validation.

Analytical method validation is always for a specific purpose and there can be no general rule. Regarding specificity (sometimes called selectivity, a close but not exact synonym), I've seen the two extremes:

Firstly in pharmaceutical quality control, you know what you're analysing because there's a separate identity test. This often has a spectral fingerprint but historically was designed in the negative sense, simply to exclude other compounds likely to be encountered. Validation of the specificity of a subsequent chromatographic (for example) analysis is largely a matter of demonstrating that known impurities and degradation products would be detected.

Secondly, when determining a new metabolite or validating a new method for an old one, it's absolutely vital to prove that you're measuring the target compound and nothing else. This is a totally different game. The best guideline I've seen is by the "metabolomics" people DOI: 10.1007/s11306-007-0082-2 (2007). This doesn't really target the analytical community: a book needs to be written though that won't be by me.

To emphasise the importance of getting this right, I can cite some articles I've written about nonsensical claims dating back three decades and still going strong, that the arrow poison ouabain is a mammalian hormone:

DOI: 10.13140/RG.2.2.24929.84323

Incidentally, that's a reminder that cross contamination can be an important validation criterion; ouabain happens to be used, in almost industrial quantities, as a pharmacological tool in the laboratories concerned...

Finally, there are likely to be other long-running biomedical stories that have been running for along time without adequate challenge regarding specificity. One of these may be about claims that the hormone melatonin is produced by organs other than the pineal gland. This may or may not have something to do with the disquieting acceptance and promotion of melatonin as an ordinary food supplement. I'm not aware of any critical review of this subject.

There are no ALTERNATIVES to method validation. You must meet all the parameters of ICH Q2 (for pharmaceuticals). However, AES is very sensitive! Your concentration is at least 10X too high! Perform a dilution to bring your concentration down to at least 1 to 0.1 ppm.

But in selectivity mention taling factor as per ICH guideline for method validation of HPTLC

Then how you decide your peak shape good in which basis?

Without taling factor u can't decide either peak is good or not

NO singal paper of HPTLC not mention taling factor and resolution if combination

Also check please the following good link: http://www.cskb.cz/res/file/dokumenty/validation-and-verification-of-analytical-methods.pdf

You can do it with Prism Graphpad

Yes, I think so, it is necessary to validate LN2 tank 1month then -80 degrees celcius 1 week then LN2 tank.

Thanks dear Fernando Aguilar-Galindo for the answer.

Method validation is a good practice. It is like a positive control and any control experiment is better than no control. However, the conclusion from a positive control can not always be extended to the test system under investigation. Here is an example with molecular docking. In molecular docking, you can take a known ligand-bound crystal structure and then do a self-docking to check if the docking method can reproduce the binding pose of the ligand as observed in the crystal structure (within a reasonable limit of deviation). If the docking method can reproduce the binding pose of the ligand as observed in the crystal structure, you can say that the method is valid and is working for your system. Then you take an unknown ligand for docking and follow the same protocol and find a binding pose with a comparable or maybe even better binding energy. But this does not guarantee that the result is not a false positive. Afterall, it is just a simulation. The only way to validate the docking result is to show at least the binding is actually there. A simple spectroscopy or calorimetry can confirm if there is actually any binding. When there is this experimental confirmation, you can conclude from the in silico result that the simulation did not produce a false positive; the binding is there and the computed binding pose is the best possible binding pose. The same goes for the MD simulation. Now coming to your question, if you try to reproduce a published result keeping all the conditions similar (software, parameters etc.), you should get a similar result. But that is not method validation. Method validation should also be tied to experimental observation. Say you take a solution NMR structure of a DNA fragment from a nucleotide database and then run MD following the protocol that you want to use for your experiment and see if the structure remains intact. If the equilibrium structure remains the same as the input, you can say that the method worked for such a system (a known DNA fragment from database) and if your system is similar (e.g. a DNA fragment with different length and composition), it probably should work for your system as well.

Scatterometry of sub-wavelength periodic structures is a highly-developed technique that for the most part relies on zero'th order diffraction. At Zygo we have in past years developed interferometric scatterometers for some specialized applications; but there are many other designs. Just web search "scatterometry semiconductor" to get started.

It depends on the analytes, on the matrices and how the validation study was carried out.

In general the selection of these QCSs is correct and it is possible to choose calibration standard levels as QC levels.

From Guide to Quality in Analytical Chemistry - EURACHEM:

"Standards or materials similar to those used for calibration, placed at intervals in an analytical batch, enable checks to be made that the response of the analytical process to the analyte is stable. It is the responsibility of the laboratory management to set and justify an appropriate level of QC, based on risk assessment, taking into account the reliability of the method, the criticality of the work, and the feasibility of repeating the analysis if the QC sample result is unacceptable."

Here is the link of full text:

Also, consulting other guidelines (e.g. ICH, SANTE, IUPAC) is a good suggestion.

KRs

To calculate the reproducibility, it is sufficient to take one peak. It should be intense, but not exceed the maximum recorded value of the optical density. The suitable peak will be determined by the film thickness and spectral resolution, it can be 725 (720 + 730) or 1467.

Dear Bongisiwe Zozo

Headspace gas chromatography (HS-GC) and reverse phase high-performance liquid chromatography (RP-HPLC) are the two main methods used for measurement of acetaldehyde. The HPLC method entails labeling with a derivatization reagent for stabilization and selective determination of acetaldehyde.

For more info, have a look:

1. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4609974/

2. https://www.researchgate.net/profile/Cho-Il-Kim/publication/283049295_Validation_and_Determination_of_the_Contents_of_Acetaldehyde_and_Formaldehyde_in_Foods/inline/jsViewer/56df7a2508aec4b3333b7621?inViewer=1&pdfJsDownload=1&origin=publication_detail&previewAsPdf=false

3. https://www.megazyme.com/documents/Validation_Report/K-ACHYD_Validation_Report.pdf

4. https://www.ijpsonline.com/articles/determination-of-acetaldehyde-content-in-candesartan-cilexetil-by-hplc-3523.html

Kind Regards

I hope its meet your needs

Also check http://www.physiology.org/doi/pdf/10.1152/ajplegacy.1955.182.3.428

YES!!: Of course;Why not??

Via thermodynamic analysis, concentration profiles can be determined based on Gibb's free energy minimisation or equilibrium (stoichiometric) balances at specified conditions (temperature, pressures)

Likewise, adiabatic flame temperature calculations involving energy balances can be carried out to evaluate temperature profiles on basis of variant inlet oxidant, fuel temperatures and air enrichment ratios.

Engineering analysis of thermal/chemical reacting/non-reacting systems can result in values for metrics such as thermal efficiency, chemical efficiency and so on and so forth that can also be calculated from the simulation results and compared

Ezekiel Yorke I believe you partly answered your question in the last sentence of the first paragraph-"for cost-effectiveness and resource-saving".

Early in my career I encountered Anne Anastasi's book "Psychological Testing". I was struck, and have remained impressed all my life, by her thought that a "questionnaire is a sample of behaviour". That is, a questionnaire is a *sample* from which the analyst makes predictions about the *ensemble* of behaviours the respondent may be said to exhibit. It makes complete sense therefore to investigate the ability of a questionnaire to allow the prediction to be made with validity (does the sample really measure the ensemble of interest?) and reliability (does the sample fluctuate at random?) Using unvalidated questionnaires is, as I used to drone to my dear students till they knew this lesson, pseudo-science and snake-oil. As in all inferential statistics, the actual *data* (in the original question, "the scores") is an accident that has just taken place. What the investigator wants to know is: what does the accident tell us ("the questionnaire")? Welcome to the wonderful world of Psychometrics!

Bergonie-Tribondeau's law based on experimental observations in rat testes applies at cellular levels (albeit with various exceptions such as lymphocytes, some gonadal cells, AT cells, and SV40-transformed cells), but less generally at tissue or organism levels. Younger people are not necessarily radiosensitive than older people where dependence of radiosensitivity on age (e.g., age at exposure, attained age) greatly differs by endpoints (health outcomes). For instance, children are more radiosensitive than adults for about 25% of cancers (e.g, leukemia, thyroid, skin, breast and brain cancers), as detailed in

Hello, the certain group of microorganisms

has cotrimixazole degradation activity such as Bacillus subtilis, Pseudomonas aeruginosa, Pseudomonas putida, Rhodococcus equi, Rhodococcus erythropolis, Rhodococcus rhodocrous, and Rhodococcus zopfii.

The most active is Rhodococcus equi.

If you want to validate cotrimoxazole tablet look for Pharmacopoeia method.

Thank you Dear Amaury Borges Miranda for your comment.

In our laboratory routine, we always work with periodically calibrated equipment. Protocol validation and optimization steps are always necessary. Mainly because sometimes we are unable to buy reagents of the same brand that we often use (due to lack of stock). So whenever there is a change in the PCR reagents, we have to optimize the protocols and check the amplification fidelity of the gene with those new reagents.

Dear Kim,

Are different parameters:

This is a measure of the difference between the expectation of the test result and the accepted reference value due to systematic method and laboratory error. It is usually expressed as a percentage. Accuracy and precision together determine the total error of the analysis. Accuracy is ideally determined using Certified Reference Materials (CRMs), if available, reference methods, collaborative studies or by comparison with other methods;

The recovery of an analyte in an assay is the detector response obtained from an amount of the analyte added to and extracted from the matrix, compared to the detector response for the true concentration of the pure authentic standard (seized materials). It may also be understood as the percentage of the drug, metabolite, or internal standard originally in the specimen that reaches the end of the procedure.

Best

JSCamara

Thank you @ Bojidarka B. Ivanova for the inputs and sharing of useful references.

Mr. Baroudi,

I would like to thank you for the wishes.

Happy New 2021 Year for you, as well.

Successful work!

Dear Jinadasa

Thank you for your information. I will check and update you.

Mark is pointing you in the right direction. This is a classic MSA opportunity the only twist is that you are using it as a Go/No Go gage which will require an Attribute Agreement Analysis and achieve an >80% agreement level. Although, if you can get continuous (or variable) data from both test systems then it would be prudent to go ahead and perform a Gage R&R (10-parts X 2-systems X 3-replicates) to better understand the repeatability and reproducibility of the two systems in which the two systems are the two appraisers. This would give the best data in the shortest amount of time. The question you would need to answer is what level of error will you accept; <10%?, <20%? or <30%? Oh yeah, before you do any of these analyses please make sure both systems are appropriately calibrated and in good working order.

Longer term conducting a chi-squared on your results may make sense if you believe they are sampling from the same population over an appropriate number of samples.

Dear Mahmudul, listed parameters are sufficent for the establishment of the method (include specificity) but if you need to submit this method for regulatory agencies or adopt for the compendial monograph, then it require full validation as per ICH Q2R1 guidelines

Yes, if by one method you analysing several compounds.

Hi, I think it is possible, but at first you should determine relative conversion factor.

http://www.tiprpress.com/chmen/article/abstract/chm201701006?st=article_issue

Every lab has its own protocols, which often depend on the type of sample being analyzed. Some labs do these analyses more than others and have honed their techniques a bit better. Fiehn's lab at UC Davis is pretty good on Metabolomics from plants. Here are some general rules to remember:

1. ALL sample preparation methods bias the results. In your case you will undoubtedly be precipitating the blood proteins after centrifuging out the cells. The age of the sample and any previous free-thawing will lyse some of the cells spilling their contents into the plasma. Even differences in the blood collection vials and manufacturers can influence the results you get. I once saw a study where they collected the control samples with one vacutainer and the disease samples with another (two different hospitals) and the 3 month LC-MS study found that lactic acid was the principal difference, which was included by the mfg in tube 2. Talk about a waste of time and resources.

2. The LC conditions will similarly bias the results. Not everything will elute from the column and the column packing will change over time. Some materials won't bind and will elute at the flow front in a jumbled mess. Elution time is one marker for identity and sample alignment.

3. The MS mode (+ ion or - ion) will suppress seeing 30-50% of the metabolites in the sample. Furthermore, some molecules won't ionize by ESI at all and require other modes. So never think that you are seeing everything. Also differences in salt compositions might cause a peak to jump from one mas to another mass because of adduct differences.

4. MS is NOT quantitative unless you are spiking with stable isotopes. The best you can hope for is ratiometric comparisons between samples and metabolites. Scaling all peaks to a common reference peak is a way to deal with this issue. In general, you are looking for 1-log differences in relative abundances to identify a difference. Ionization efficiencies simply vary by an order of magnitude over time and runs and more between different molecules.

5. An open method is just that. It is designed to flag potential real differences between samples, that deserve follow-up with additional hypothesis-driven experiments. It is not a validation study. The issue here is basic chemometrics. You need a minimum of 3 independent samples per feature analyzed to actually make a statistical call. In a typical metabolomics study, you might have 1000 features, which means you can't make any real statistical call on any feature without running 3000 patient samples. No one can afford that, so you use what you can afford to run to try to identify leads that deserve independent followup. You have to remember that the lower the sample to feature ratio, the more false leads that will be generated on an exponential or hyperbolic scale.

6. Spend the time to really understand the math in each step of your analysis. Blindly running the software at its default settings will always lead you astray. Until you understand exactly how each step is working, you can't begin to understand what biases the software is introducing to the analysis. This starts at centroiding and doesn't end until the PCA analysis at the end. I always run in profile mode (not centroid mode) because I don't trust that the MS Mfg has implemented his code correctly for complex samples. In fact, I've proved that on pretty much every MS. Then you have 'hidden' features in software like OpenMS and MZmine that when they flag a missing peak in the sample, they go back to the profile mode spectrum to find the highest point in the neighboring baseline and report that instead. Sorry, just not statistically-valid.

It was some time ago, thank you for your advise. I agree with you.

Regards,

Tachi

Hello Daniel J F Santos ,

Yes it will. You can see the below document.

S. Wagner, "Environmental Fate of Imidacloprid," Sacramento, CA, 2016.

Furthermore, the coupling of UV light increased the degradation of Imidacloprid by ozonation. You can check my paper if you would like to.

Best.

Busra S. Baghirzade

Nice Dear Wiem Khlif

dear dr Jaydip Datta thank you for link i surf it there is such answers i get from

Is there any guidelines for in process hold time study of semi-solid dosage form

Hello, see paper attached which describe all requirements in PCR method validation.

Good luck!

The reference that Noman Mohammad gave you (Lincoln and Guba 1985, Naturalistic Inquiry) is the best known source on validity in qualitative research. Note, however, that if you are working with mixed methods, reviewers will be much more likely to care about reliability and validity with regard to the quantitative rather than the qualitative portions of the overall project.

To clarify the term validation shortly: in the field of antibodies, validation includes mainly specificity and sensitivity in a certain application. To show specifity, genetic methods like specific knock-down of the target antigen in a cell line would be the method of choice. The sensitivity is important to show that your antibody works in a specific assay e.g. Western Blot, FACS, IHC etc. You could have the most specific antibody for an antigen but that doesn't guarantee that it will work for both Western blot and IHC for a specific cell line, for example. That part of the validation is important and needs to be done for the specific application.

Note that using recombinant protein as a control is not 100% reliable in order to verify that your antibody will work for your cell line or lysate. Post-translational modifications, different isoforms in different cell lines or low expression levels could give negative or unclear results even if your recombinant protein is detected, especially when it was produced in E.coli, you might have differences compared to the protein from eukaryotic cells.

You asked about the validity and reliability. These are the critical points. If you chose your subject from very dissimilar groups and settings there is a greater chance to get reliable data as well as external validity which is important for researcher who want to repeat the experiment. Check this out if it gives you advice: https://pdfs.semanticscholar.org/6462/54a1698a66aacb6eceb66c126309c3311997.pdf

Hello, you can keep as long as you need for further analysis. Before storage perform at least 6 injection in repeatability conditions and calculate standard deviation (SD) for creation of quality control chart with 2xSD limits than after storage each week perform 1 duplicate analysis. If you receive result within range it mean your sample still stabile.

Good luck!

Hello if you have method X which already validated and all requirements and performance criteria for the method are fulfilled, so you can use parameters from method X to validate method Y. If parameters from Y method equal to X based on it you can say that Y also validated.

These forums do not provide a means to address such a broad question as posed. The basics of HPLC as a formal analytical technique and method validation basics fill volumes of books and journals. Read the books to research the basics. As HPLC method development and HPLC method validation take many years in the workplace to acquire a basic level of proficiency, I suggest that after researching the topic, you hire a contract laboratory to perform the work for you OR work with a local laboratory whom has experienced liquid chromatographers with a demonstrated history of doing the same for compounds of interest.

Additionally, you may wish to take some time and begin to research the basic of liquid chromatography [e.g. The text, "Introduction to Modern Liquid Chromatography" by: Lloyd R. Snyder, Joseph J. Kirkland, John W. Dolan is excellent] , HPLC methods, the basic concepts and fundamentals of analysis. BTW: Avoid sales oriented HPLC websites such as chromforum or chromacademy which have many inexperienced users who may misdirect you. Additionally, if you wish to be involved in the actual analysis work, then you will need to obtain formal training on the specific HPLC system and software that will be used for the work (Many types of HPLC systems and each one requires training in how to use it).

The decisions involved in combining any two methods within mixed methods depend on not only the purposes that each separate method will serve but also the ways in which you plan to integrate what you learn from each method.

One possible mixed method approach that you might consider is a sequential Explanatory design. (QUANT --> qual). In this case the qualitative method serves as a follow-up study to help explain the results of the original quantitative study. In particular, the role of the qual study is often to explain how and why the quant study produced the results that it did.

First, let me suggest that one might consider not spiking at the MDL even if one could estimate it correctly. Errors can be better estimated close to but not at the MDL. Spike to high and your MDL is too large, spike to low and the measures have so much error again your MDLs will not make sense with observation.

You are correct that an estimate of the population standard deviation is divided by the square root of n. However, dividing by square root of n is used for arriving an estimate of the population standard error (s estimate of u). In this case, one is only interested in using the standard deviation of your measures to arrive at a good estimate of the MDL witthout spiking at it directly. Of course, this is acheived by the variability (sd) of your measures being multiplied by the student's t value (n-1) at 99% probability (typically chosen). I prefer to define a limit of determination (LOD) which is defined as the least amount that can be identified and measured with a given predefined accuracy. The abreviation "LOD" is unfortunately to easily confused with limit of detection which it is not. So were rarely speak of it that way. EU has the same difficulty with this term.

The non-specific bands detected in WB, often minor, are usually artifacts related to the use of cell lysates (or impure proteins) containing cytochrome C or IgG receptors ... The differential between positive cells and cells negatives makes it possible to guarantee the specificity of marking. In publications one is careful not to show these artifacts in order to avoid having to give explanations. As for the blocking of the labeling by the specific antibody, it is imperative to put very large quantities of target proteins to obtain a specific inhibition: of the order of 10 to 100 times the amount of antibody applied. On the other hand, the antibody having a high affinity for the whole protein as for the synthetic peptide, it is preferable to use the entire protein when it is possessed.

Dear Ajay,

This an interesting question and I am afraid that the advice you got form the previous posters is only of very limited value. It is so because of three facts:

1) Protein folding as a science problem (a question) is an ill-defined problem. Ill-defined means that does not have well formulated question nor the answer to this question has very well worked out (robust) methodology. You are welcome to ask why I think so, but that is a separate issue requiring a much longer discussion.

2) The available ab initio methods are all using at least some heuristic arguments or methods what makes them self confirmatory not exploratory. So even the best methods as Rosetta despite all the hoopla is incapable answering a significant number of problems with required accuracy.

3) The existing validation methods are build on the existing databases what assumes minimal errors contained in them. This is contrary to my experience which shows that almost 75% of proteins retrieved form PDB, in my hands, contain small or large errors. Removal of this errors is almost impossible because of resistance of the community at large. So before using them in my projects I reach the experimental limitations myself by redefining them or remodeling them.

The only real validation comes form experimental methods that produce their own errors. So the problem is complex and requires iterative (almost Kuhnan approach) of hypothesis forming and testing by experiment. So use all the methods cited by the previous posters but not be surprised if some experimental results come diverging form expectations. In heuristic methods the closer the template to the target the better but in my hands two cases of very close similarities resulted in diverging answers. This is so because above mentioned ill-defined character of the initial question.

Good Luck.

Bog

Been done! See attached;

Using certified standards to analyze m / z for both the precursor ion and its fragments.

As above mentioned answers, all are helping; if questionaire has aready been used in some study and it has been passed through all sorts of validity process, and your research question is similar to the previous one then you can easily use that questionare, but if it is new one and you have developed it, then face validity(show ypour questionaire to some expert in the field or your mentor, they can check for face and content validity) Dont get confused, these things seem complex apparently but they are actually not. Then you can piolt that questionaire before conducting actual study (piloting can be done on the 10% of sample in accordance with your actual sample; but sometimes five percent sample is also used, it is stated differently in many books, but for practical purpose, piloting is good thing as sometimes what questions we have posed are comprehended by the participants differently; and if discrepancy exists what we are asking and others are answering, then questions can be modified). Peer review of your questionaire can also help; purpose of all this practice is simple as our questionaire must measure what it needs to be measured, oranges comparison must be with oranges and this process is simple but lengthy and needs careful, common sense. Expert responses on questionaire can also be halpful to see inter rater and intrarater reliability.....Good luck!

Hello, The most commonly used method is statistical pattern recognition is the Bayes ”plug-in” classifier. The usual kernal that is plugged in is the multivariate Gaussian distribution is attached.

Good luck!

Hello see attached info may be useful for you.

Good luck!

Epistemology reveals that the human mind is continually in a process of becoming.

Although 24HR recall is subjective and a single administration of 24 hr recall is unable to account for day-to-day variation, it is still a valid measure of dietary assessment and can be used for epidemiological purposes.

If a more accurate assessment is required it can be combined with FFQ.

I think to information literacy of media agency in share information, all regards...

Hello, send your questionnaire to the competent peoples which have experience in same area for comments and remarks.

Try ROC curves using data or examples presented to your researchers.

These data / examples should be (blinded for your researchers) evaluated by an expert or given a consensus result by a group of experts.

Salvatore S. Mangiafico Yes.

Hello, see attached paper may be helpful for you.

Good luck!

Common P-EIS^[1] is non destructive^[2], if you performed a measurement under the conditions:

1) a low V_ac(:=1 to 5, mV) amplitudes,

2) a low over-potential stress V_dc(=OCP), and

3) short (stress) measurement time (<near, 1h, or less).

you might find the following helpful

https://jech.bmj.com/content/jech/56/3/177.full.pdf

The Gilens and Page Measure of Political Inequality

However, it's more likely that you will find measures if you define your variables of interest more as it's unlikely that one measure can capture all aspects of power and inequality.

Dear Fabiola,

in general you can use different approaches. First of all you could show and validate that external calibration is applicable. Some researchers also use surrogate matrices like e.g. saline buffers + BSA to create an artificial matrix or you can also do charcoal stripping of your sample to remove your analyte. However, all these approaches will not let you determine the true matrix effect. I prefer the surrogate calibrant/analyte approach. Here you prepare calibrants and QCs by spiking isotopic labelled analogues of your analaytes into the real matrix and using these signals for quantification of your target analytes.

Here´s also some literature on the topic

Glad you figured out the hardest part of HILIC, which is the equilibration. It takes 19 forevers to equilibrate properly, and 19 forevers to re-equilibrate after each run. There are better, more stable methods for sugars than HILIC. Take a look at mixed mode columns. I think you will like them.

There are no appropriate texts! ICH validation guidelines apply to pharmaceuticals and biopharmaceuticals. With small molecules it has been relatively easy to validate a method but stem cells will be very hard. Is a recovery of 50% for a fluorescent antigen technique appropriate?

The 'validation biologist' will have to demonstrate that the method is consistent and reliable. This will mean setting the appropriate and relevant acceptance criteria or specifications.

Thank you Richard Odemer Bruce Neagle

Thank you all.

@ Mahmoud H El-Maghrabey ,

ICH is a good guideline , but it does not mention any criteria.

Dear Urvesh,

You may refer to FDA Bioanalytical Method Validation guidance (May 2018): „The sponsor should optimize the recovery of the analyte to ensure that the extraction is efficient and reproducible. Recovery need not be 100 percent, but the extent of the recovery of an analyte and of the ISs should be consistent and reproducible.” Summing up, if the recovery is constant across your calibration range (within +/- 15%), then your method will produce reliable results.

Best regards,

Piotr

If you are performing Method Precision, then it should be 2.0% as per ICH Q2(R1).

Accuracy and Specificity (placebo interference) can be done by flask method also.

Proof by contradiction is valid only under certain conditions. The main conditions are:

- The problem can be described as a set of (usually two) mutually exclusive propositions;

- These cases are demonstrably exhaustive, in the sense that no other possible proposition exists.

Under these circumstances, if all but one of the cases are proven to be false, the remaining case must be true.

It is useful when the proposition of interest is hard to prove, but the contradictory proposition(s) is/are easy to disprove. These conditions do, of course, apply well to many mathematical problems.

For example the simplest proof that the square root of two is irrational is a proof by contradiction. We can state the problem as two mutually exclusive cases; A: sqrt(2) is irrational; or B: sqrt(2) is rational. And this is an exhuastive set; there is no other possibility. It is quite hard to show that A is true directly; but it is quite easy to show that assuming B generates a contradiction and must therefore be false. (See https://en.wikipedia.org/wiki/Proof_by_contradiction#Irrationality_of_the_square_root_of_2).

In the 'natural' world, this still works when these basic conditions hold. Finding one black swan (or a blue one) is essentially a proof-by-contradiction that "All swans are not white". (observation of a single black swan conclusively contradicts the only possible alternative proposition, that "all swans _are_ white").

But the natural world is often messier than this. First, we are often much more interested in partial generalisations (eg "_most_ swans are white") and that is impossible to prove or disprove conclusively without exhaustive counting (though statistical inference allows the statement to be rejected as improbable after suitable sampling exercises). Further, there are often many alternatives of interest, we cannot always disprove all but one, and even if we can, we cannot rule out the possibility of another, unknown case. Proving something is not a cat does not prove that it is a dog. It may be a mouse. And if we also prove it is not a mouse, there are plenty more furry animals to work through, and even after we have gone through all the known furry animals, we may be still looking at a new, previously unknown, species.

So the principle always works if the basic conditions above hold; but it is not valid if they don't. And in the natural world, if those conditions _do_ hold, we're either very lucky, or we have made some improbably sweeping generalisation that is trivially refutable.

No, Robustness and Ruggedness studies were not require for UV Spectroscopic method.

If your HPLC procedure is just optimized, you need to perform the following experiments:

- linearity

- LOD and LOQ

- precision and trueness

- recovery

- stability (short and long-term)

For each point you need to perform 6 independent experiments.

For linearity are necessary at least 5 non zero points at different concentration levels.

Additionally for precision and trueness you need to analyse 6 times 3 QC levels (low, medium, and high and the concentrations need to be different respect the concentration levels used in linearity) obtained by spiking (with known amount) blank matrix.

All these procedures are deeply discussed at the following links: