Ming Gu

Recent advances in the development of new instruments and innovative technology have facilitated the elemental composition determination for unknown identification and compound confirmation significantly. Routinely delivering high resolution to separate ions of interest from background signals and high mass accuracy of 5 ppm or better, many types of new instruments including time-of-flight (TOF), orbital trap, and Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometry (MS) systems have been recognized as powerful tools for empirical formula determination. Their capability of formula identification has been further enhanced by utilizing the isotope information (1–4). Of particular importance is the innovative peak-shape calibration that results in mathematically well-defined and symmetrical mass spectral peak shape, which allows exact isotope modeling to improve formula identification performance drastically. Even for an orbital trap system, up to 99% of false positives can be eliminated with the high spectral accuracy achievable, as demonstrated in a recent publication (2). The same technology also made it possible for accurate mass measurements and formula identification with a unit mass resolution quadrupole mass spectrometer (5).   

Regardless of this impressive progress in both high mass accuracy and high spectral accuracy, unique formula identification for true unknowns remains a formidable task. This is particularly true, for example, in the identification of natural products and impurities of pharmaceuticals, where there is little knowledge, if any, of the unknowns in terms of possible elements included and their lower and upper limits. To meet the challenge, new approaches known as "seven golden rules to formula identification" were proposed by Kind and Oliver (6,7). Their pioneering work on filtering false positive formulas during elemental composition determination was based upon statistic investigation of the compounds in databases and the first principles of chemistry. Developed from comprehensive analysis of 68,237 unique compounds from both PubChem and Dictionary of Nature Products databases, these seven golden rules can be classified into three categories:  					

Chemical elements related ratios (rules 4 and 5), probabilities (rule 6), and their upper limits (rule 1)

Chemistry principles of Lewis and Senior rules (rule 2) and isotope patterns (rule 3)

Chemical functional group specific rule for electron ionization MS (rule 7)

While rigorous validations have been done by the statistics on a large formula library and computer simulated spectra, these rules were tested only by limited, experimentally acquired high-resolution data. In this work, the applications of rules 1, 4, and 5 to unknown compound identification using both unit mass resolution and high-resolution data acquired from single-quadrupole, TOF, and orbital trap MS systems will be focused on. Because exact isotope modeling for formula determination (rule 3) has been implemented successfully through software (MassWorks, Cerno Bioscience, Danbury, Connecticut), this investigation will demonstrate which of the three rules will further enhance the formula determination after CLIPS or sCLIPS formula search.  

Data Acquisition of Unknowns and Calibration Standards: 

All the data acquisition from both unit mass resolution and high resolution instruments was performed in a profile mode with the threshold set to zero where applicable to ensure accurate isotope measurement. The gas chromatographic (GC)–MS acquisition of tetramethylenedisulfotetramine was performed in a GC–MSD system (Agilent Technologies, Santa Clara, California) in raw scan mode at a scan speed 2^2 (A/D samples = 4) over a mass range of 40–550. After the compound elution at 8.7 min, PFTBA ions were collected for 1 min to be used as calibration standards.

In liquid chromatographic (LC)–MS analysis, two separate runs were acquired on an SQD system (Waters Corporation, Milford, Massachusetts). One was for the mixture of atenolol, imipramine hydrochloride, enalapril, and Tyr-Tyr-Tyr, which were calibrated by a set of calibration standards consisting of busprione, reserpine, and erythromycin. The other was for the degradation products of simvastatin (8). Because the 

 value of simvastatin is close enough to the 

 values of its degradation products, simvastatin and its sodium and potassium adducts were conveniently used as calibration standards for the degradants. Additional high resolution LC–MS analyses were performed for probenecid and erythromycin on an Orbitrap MS system (Thermo Fisher Scientific, Madison, Wisconsin) and for Tyr-Tyr-Tyr on an LC–TOF system.  

All unit mass resolution data were calibrated by their respective calibration standards. This calibration is the key for unit mass resolution quadrupole instruments to achieve high mass accuracy and exact isotope modeling for CLIPS formula determination. Rigorous description on the calibration was reported previously (9,10). Briefly, the calibration takes both monoisotope 

 values and isotope distribution of calibration standards into consideration. In GC–MS, for example, a set of PFTBA ions, ranged from 

 68.9 to 501.9, were selected to generate the calibration. Both their 

 values and isotope profiles were calibrated according to the exact mass and theoretically calculated isotope distribution provided by the molecular formula of C

 values were corrected and also their peak shape was calibrated to a symmetrical and mathematically well-defined function. This calibration was then applied to unknown data to achieve high mass accuracy and exact isotope modeling for formula determination. For high-resolution data acquired from orbital trap or LC–TOF MS, the calibration was performed on the peak shape only, because high mass accuracy presumably can be obtained from these high-resolution data already.   

Mass tolerance of 15 mDa and 2 mDa were set for the elemental composition determination for unit mass resolution and high resolution data respectively. With no particular restriction, double bond equivalent range was selected from —1 to 50. For molecular ions in GC–MS, an odd electron state was selected, while an even electron state was used for LC–MS data. The upper limits of elements can be set through one of the two ways available in the software. One way is the theoretical maximum, given by the 

 value of the unknown divided by atomic weight for each element. For example, the theoretical possible upper limit of carbon at 

 240 is 240/12 = 20. The alternative limits are determined by the largest possible chemical elements at various 

 from compound library statistics and proper interpolation between available 

Formula searches for unknowns were performed first based uponly on high mass accuracy and high spectral accuracy and followed by the application of these three rules. These constraints were applied separately one after another so both their individual and cumulative effect on filtering out false positive formulas can be evaluated.  

Formula identification of the true unknowns requires generous open search conditions so that the upper and lower limits of possible elements should be set to theoretically, the maximum (

 divided by atomic mass) and zero, respectively. As a result, the number of possible formulas is typically quite large, even though only the most common elements C/H/N/O/S are considered. This problem is further aggravated as the 

 value of unknown compounds increases beyond 500 Da. For example, at 2 ppm mass tolerance, there are 12 possible candidates for buspirone [M+H]

 386, but the number of possible formulas increases to 87 for erythromycin [M-H]

 732. Many of the formula candidates for 

 386 and 732 can violate fundamental chemistry rules or be statistically improbable.  

As demonstrated by the formula search for tetramethylenedisulfotetramine, with five possible elements of C, H, N, O, and S and nine possible elements of C, H, N, O, S, F, P, Cl, and Br (Table I), these false positives can be removed based upon rules 1, 4, and 5. It was found that rule 1 was the most effective one to eliminate wrong formulas, resulting in 42% and 62% reduction from the five and nine elements searches, respectively, while only a 7% and 12% reduction were made accordingly by rules 4 and 5 combined. These results indicate there are more formulas having elements exceeding the maximum number set by rule 1 than those having incorrect element ratios determined by rules 4 and 5. The effectiveness of rule 1 was due to its significantly decreased upper limits for elements N and S in this example. Once rule 1 is applied, the maximum number of N and S was reduced from the theoretically allowable 17 and 7 to 5 and 3, respectively, leading to the elimination of any formula containing more than five N atoms or three S atoms.  

Even though the quantities of false positives filtered out by the rules are an important indication of the overall efficiency for formula reduction, it is more important to examine what spectral accuracy those eliminated formulas have and whether any of the elimination occurs within the top three hits, in which the correct formula usually appears with about 90% probabilities. Indeed, in each of the five-element and nine-element searches shown in Table I, two formulas with spectral accuracy better than 98.5% (Table II) and one formula with spectral accuracy at 98.9% (data not shown) were removed from the top three hits, respectively. Because of their high spectral accuracy and high ranking, these false positives could hardly be distinguished from the correct formula. Their removal by rule 1 from the top three hits leads to more confident compound identification. 

The Reduction of Formulas on Top Three Hits: 

With a focus on formula reduction on the top three hits, additional formula search was conducted for a total of seven compounds. They were acquired from either single-quadrupole or high-resolution MS, covering mass range from 240 to 732 Da. Summarized in Table II,  out of these eight formula determinations, five had two incorrect formulas filtered out of the top three hits and three had the first hit removed as incorrect formulas. To estimate the impact provided by rule 1 to the differentiation between the first hit and the second hit, the spectral accuracy (SA) difference ΔSA was calculated before and after rule 1 was applied. All ΔSA values from the search with rule 1 show significant increase. Five out of eight ΔSA values increased from 0.3–1.1% to over 1.4%. This difference appears to be too small  to be meaningful by conventional wisdom, but it is statistically significant for high-confidence unknown identification due to the exact isotope modeling enabled by peak shape calibration technology. As the best example, the initial search for probenecid resulted in C

S (correct formula) as the first and second hits, having spectral accuracy of 97.0% and 96.3% respectively. After the search with rule 1 was enabled, both the first hit and third hit from the initial top three were removed and the correct formula C

S became the number one hit. As shown clearly in Figure 1, the ΔSA value for this compound increased from 0.7% initially to 4.1%, largely due to the absence of one S atom in the second formula C

Formula determination for true  unknowns can be facilitated by the heuristic rules. The rule on the upper limits of elements (rule 1) was found to be the most effective among the three rules. This rule helps to filter out the majority of false positives. More importantly, it eliminates incorrect formulas from the top three hits obtained by exact isotope modeling. Such reduction of the false positives with high spectral accuracy from the top three hits significantly boosts the confidence of formula determination. With the added capability provided by these heuristic formula rules, the software described here is delivering the most comprehensive and powerful formula identification tool for mass spectrometrists.  

 is with Cerno Bioscience, Danbury, Connecticut. 

(1) K. Hobby, R.T. Gallagher, P. Caldwell, and I.D. Wilson, 

(2) J.C.L. Erve, M. Gu, Y. Wang, W. DeMaio, and R.E. Talaat,

(3) K. Hobby,Waters Application notes, Lit Code Number, 720001345EN 

(4) C. Koester, (Bruker Daltonik GmbH, Germany). Application: GB, GB, 1999. 

(5) M. Gu, Y. Wang, XG. Zhao, and ZM. Gu, 

(8) M. Gu, Cerno Bioscience, Application Notes, Number 105 (2008)

(9) Y. Wang, U.S. Patent No. 6,983,213 (2006). 

"Current Trends in Mass Spectrometry," 38–43, November 2007. 

Articles

A new approach to enhancing the performance of formula identification of true unknowns beyond high mass and spectral accuracy was evaluated. Three heuristic rules on upper limits and ratios of elements were tested for their effectiveness in filtering out false positive formulas with both high- and low-resolution mass spectrometry data. The rule on elements' upper limits was found to be the most effective one in eliminating incorrect formulas.

Enhancing Mass Spectral Formula Determination by Heuristic Rules

With electron ionization (EI) used in most gas chromatography–mass spectrometry (GC–MS) applications, the molecular ion typically is broken apart into multiple fragment ions during the EI process. These multiple EI fragments complement each other to enhance the compound identification capability greatly through the use of MS library search (for example, NIST MS library search), in spite of the fact that the mass-to-charge ratio (

 values in the library are accurate only to about ×0.5 Da. Consequently, EI library search has become the method of choice for GC–MS compound identification. While this works well for many routine applications, ambiguity in compound identification can occur in the case in which similarly high search scores are obtained, corresponding to multiple compounds. In demanding applications such as the identification of impurities, degradants, and active ingredients in pharmaceutical, chemical, food, flavor, and fragrances industries, these unknown compounds might not even be in a database to start with. Consequently, these applications would require accurate mass measurement from a higher-resolution MS system such as GC–time of flight (TOF), which is more expensive and requires higher maintenance. The accurate mass measurement would serve either to confirm the library search results or to determine the elemental compositions of molecular ions and key fragments.  

In a separate publication, Chen and colleagues (1) reported that it is entirely feasible to perform the same highly accurate measurements on a conventional quadrupole GC–MS system, through the use of a novel calibration scheme to calibrate both 

 and mass spectral instrument line shape elaborately. Because this elaborate calibration scheme involves both mass accuracy and spectral accuracy, a question typically arises concerning the stability of such an extensive calibration and its implications for practical applications. 

This article reports on the design and results of a short-term calibration stability study in terms of both mass accuracy and spectral accuracy, and their impacts on elemental composition determination using a conventional single-quadrupole GC–MS system. 

The conventional mass spectral calibration determines only the mass (

) alone without any consideration of the instrument line shape, which can impact the mass measurement significantly, especially on a unit mass resolution instrument such as a quadrupole GC–MS system. Instead, a novel calibration approach has been developed (2,3) that not only calibrates the mass axis, but also calibrates the mass spectral instrument line shape. This is uniquely possible in MS because the theoretical peak positions are known very accurately (they are simply the masses of corresponding ions). In addition, we also know quite accurately the various isotopes and their relative abundances. This information allows the calculation of a calibration function that corrects not only the peak position, but the instrument line shape as well.  

With this new calibration technique, the actual instrument line shape, typically unknown and undefined, is calibrated into a symmetrical and mathematically defined line shape function. This can improve all downstream processing dramatically, including accurate monoisotopic mass determination to achieve high mass accuracy on a single-quadrupole GC–MS instrument. Conventionally, with just accurate mass information, the elemental composition has been determined by searching for all chemically possible formulas within a mass tolerance window from the reported accurate mass. Depending upon the inherent mass accuracy of the instrument, all formulas found within the mass tolerance must be considered. For example, for an instrument capable of a mass accuracy of 1 ppm, 34 formula candidates for a compound at 500 Da bounded by the elements C, H, N, O, S, and Cl are identified. 

It is known that every unique formula has a unique isotope distribution, and that, in theory, could be used as additional information for formula identification. Previously, however, such methods were not considered viable for formula identification because it was difficult to model the isotope patterns on any given instrument to a spectral accuracy of better than 98% (4). This has limited the use of isotope profile information to assist in reducing the number of formula candidates generated from accurate mass measurements. If, however, the spectral accuracy could be improved to better than 99%, it could be used to identify a unique formula from a list of possible formulas generated by the accurate mass measurement, or even provide confident formula identification on instruments that are not calibrated as frequently, as long as similarly high spectral accuracy can be ensured. 

With the calibrated instrument line shape function described earlier, a theoretical mass spectrum can be calculated exactly for each candidate formula from the list and compared with the measured mass spectrum after the previous calibration. A spectral accuracy metric of better than 99% can then be calculated to reflect any minor yet elementally significant difference between the two and rank the likelihood of a given formula being the correct answer. This process is called calibrated line-shape isotope profile search (CLIPS) and has been demonstrated to be more informative than mass accuracy for formula determination (1). 

Perfluorotributylamine (PFTBA) calibration standard and a sample mixture of compounds were measured on an Agilent GC–MS single-quadrupole MS 5973 system (Santa Clara, California) over a one-week period. The PFTBA standard was measured once through infusion and repeated again a week later at the end of the experimentation. The mixture sample was analyzed 22 times at different concentration levels between the two PFTBA runs, in which one blind test ion (nominal mass at 

 162) was to be identified for its elemental composition using MassWorks CLIPS (Cerno Bioscience, Danbury, Connecticut). Both the PFTBA and the sample mixture were acquired under the same MS conditions in "raw" mode (no-peak detected, sometimes called profile mode) over a mass range of 50–550 

The same calibration, created with eight known fragments from the first PFTBA run, was applied to all 22 GC–MS runs to report the accurate masses of this blind test ion followed by its elemental composition determination. This same calibration also was applied to the second PFTBA run to report the accurate masses of the same eight known fragments in order to measure the mass drift between these two PFTBA runs.  

Figure 1 shows the total ion chromatogram of the first PFTBA run and the 0.8-min retention time window used for the mass spectral averaging before the calibration. Figure 2 shows the profile mode mass spectrum before and after the calibration for the calibration ion C

+ . This demonstrates that the raw profile mode mass spectrum, although very well tuned, does not have a known or defined peak shape, whereas the calibrated mass spectrum does have a known and mathematically defined peak shape. The calibrated mass spectrum also has been shifted properly to the correct mass position given by the monoisotopic mass of this calibration ion. As a convention and without the loss of generality, the mass difference due to the gain or loss of an electron is ignored throughout this article and the exact mass corresponds to that of the neutral form of the same ion.  

This calibration was applied to the second PFTBA run acquired one week later, and the reported accurate masses for the same set of eight calibration ions are listed in Table I. We see a systematic mass shift of about +7 mDa across all of these ions. When measured in parts per million (ppm), such mass shift amounts to +12 to +54 ppm mass error, typically deemed too large for elemental composition determination through the conventional monoisotopic mass search. 

The same calibration was applied to the 22 GC–MS runs acquired within the week, and the accurate masses for the 

 162 ion are plotted in Figure 3 along with its exact mass, calculated after the identity of the blind test ion was revealed to be nicotine (C

) and its ion signal counts. There is a clear correlation between the mass error and the ion signal level, as expected and established in the literature (5). For intensities above 8000 counts, all reported accurate masses come within a few millidaltons of the exact mass. This is consistent with the overall mass shift estimate from Table I and confirms that the overall mass shift observed between the two PFTBA runs are more or less comparable to the random fluctuation of the repeated mass measurement at similar ion intensities. In other words, the overall systematic mass shift error of ~7 mDa is not statistically significant when compared with the random fluctuation of the mass measurement itself (that is, the first PFTBA calibration is indeed applicable to all 22 GC-MS runs acquired afterward without causing significant systematic mass measurement errors). 

Table I: Accurate masses of PFTBA calibration ions for data acquired on the final day 

As has been reported (6), while mass accuracy is important for formula determination, it is typically just a single number picked out of the complete isotope response of a given ion. For this reason, there are many formula candidates whose exact monoisotopic masses fall within a small mass tolerance window of a reported accurate mass, even when measuring with a high-resolution mass spectrometer with 1–2 ppm mass accuracy. Spectral accuracy, an independent metric, measures how well an actually measured mass spectrum matches that obtained from theoretical calculation. It contains much more information about the fine structures of the isotope distribution for a given ion and might be more important than mass accuracy. Furthermore, the spectral accuracy metric might be more robust than mass accuracy for the purpose of formula determination, since the mass spectral instrument line shape on which spectral accuracy depends might be less susceptible to time-related variations on a mass spectrometer. 

The unknown ion at 162 Da with its mass error of more than 50 ppm (Table I and Figure 3) offers a good example to test the relative importance and robustness of mass accuracy versus spectral accuracy. Treating the 162 Da ion as a total unknown with common elements C, H, N, O, F, and Cl included under general elemental composition search conditions (Table II), all 22 mass spectral measurements of this same ion were run through the software to determine its formula. Ranked by spectral accuracy, the formula C

 came as the first hit 19 times, as the second hit two times, and as the third hit once (Figure 4). This formula was later confirmed to be the correct formula for nicotine. The second hit and third hit all correspond to lower concentration samples (Figure 3). 

Table II: MassWorks CLIPS formula determination parameters 

Take three runs: 24–17–12, 26–18–29, and 31–11–40 as examples and overlay the theoretically calculated mass spectrum for nicotine on top of the software calibrated version. We see from Figure 5 that a good quality spectral match is attained for the lower concentration run (24-17-12) and a near perfect spectral match is attained for the higher concentration runs (26-18-29 and 31-11-40), all reflected in their corresponding spectral accuracy metric. Figure 5 offers evidence that the spectral accuracy is directly related to ion signal level (noise in data) with little or no discernable correlation to time or mass shift. This indicates that spectral accuracy is not only more informative but also far more robust for elemental composition determination. 

Figure 5 depicts a section of the mass spectrum corresponding to the 162 Da nicotine ion. This shows a mass spectral peak at 161 Da with about the same ion intensity as the nicotine ion itself. As often happens in GC–MS and tandem MS-MS experiments, there might be other ions or fragments formed during the ionization–fragmentation process that lead to such isotope response overlaps. This mass spectral interference has caused some systematic mass measurement errors for the 162 Da ion, as evidenced by the mostly negative mass measurement bias from Figure 3. This does not affect the spectral accuracy calculation because the 161 Da ion also is added to the CLIPS calculation by modeling it as an M-H ion, where M is the unknown 162 Da ion for which the elemental composition is to be determined. In theory, many more interference ions similar to the 161 Da ion could be handled, though only up to two interferences can be included at this time.  

Although the single-quadrupole GC–MS system used here has mass variation on the order of 7 mDa or more than 50 ppm at 162 Da during the one-week test period, an error large enough to cause alarms for conventional accurate mass elemental composition determination, the mass spectral line shape reflected by the spectral accuracy metric remained unchanged throughout the week, leading to consistently high confidence formula determination without recalibration. 

This short-term stability study shows that the elaborate calibration involving both mass accuracy and spectral accuracy does not render the calibration more susceptible to time-dependent variations. Furthermore, the formula determination approach maintains its robust performance through the separation of mass accuracy and spectral accuracy. While the mass accuracy fluctuated by about 7 mDa during the one-week test period, the mass spectral instrument line shape or spectral accuracy experienced no noticeable changes, allowing for robust formula determination without recalibration. This is achieved by taking advantage of the exact isotope modeling process embedded in the CLIPS algorithm, which fully exploits the fine structures in the whole isotope distribution of a given ion instead of relying on a single accurate mass reading from its monoisotopic peak. 

The authors would like to thank Ms. Alice Laures and her colleague Mr. Ian Collin from GlaxoSmithKline (Brentford, Middlesex, United Kingdom) for acquiring the GC–MS data used in this study. 

 are with Cerno Bioscience, Danbury, Connecticut. 

(3) Y. Wang, Methods for operating mass spectrometry (MS) instrument systems, US Patent 6,983,213, filed October 20, 2003. 

(4) T. Kind, BMC Bioinformatics 2006, 7, 234 (

americanlaboratory.texterity.com/al-online/200801ol/?pg=18

With electron ionization (EI) used in most gas chromatography–mass spectrometry (GC–MS) applications, the molecular ion typically is broken apart into multiple fragment ions during the EI process.

Calibration Stability for Formula Determination on a Single-Quadrupole GC–MS System

The dependence of mass accuracy (expressed as standard error σ in ppm) upon signal strength (

) has been reported in the literature (4, 5) as:   

 includes such factors as signal unit conversion to real ion counts, peak sampling interval, peak analysis, and mass determination algorithms. 

Reflecting on this equation, two important conclusions can be drawn: 					

Mass accuracy is proportional to the square root of the signal strength; and

 is dependent upon the data analysis methods outside of the physical measurement process.

The first conclusion indicates that even for low-resolution instruments, we should see a significant gain in mass accuracy for measurements with high signal-to-noise ratios (S/N). This should be particularly useful in low-resolution instruments, which tend to have increased ion transmitting efficiency. This means we should not see a significant loss in mass accuracy in instruments of low resolving power due to their intrinsically higher ion transmitting efficiency (or an effectively larger aperture, to borrow a term from optical spectroscopy).  

The second factor involves the mysterious constant 

. As defined above, it is clear that many of the parameters relating to 

 also lie outside the measurement process. This provides the opportunity to minimize the errors by way of calibration and analysis methods. Interestingly, published values for 

 rarely have been calculated for commercial instruments with the exception of the article mentioned above. By refining the analysis process, it should be possible to reduce the error for both high- and low-resolution instruments.  

 Table I. A brief summary of improvements in Mass Accuracy (MA) using HAMSCA.			

Based upon this simple analysis of errors, the authors proceeded with an in-depth analysis of the current methods of peak analysis. Current methods can be summarized as follows: 					

Run a calibration standard with multiple ions covering the mass spectral range of interest

Locate ion peaks and calculate peak centroids (peak centroiding)

Fit a calibration correction curve between the calculated peak centroids and known peak masses 

Run the sample, possibly including an internal standard

Apply the calibration correction curve to the mass axis of acquired scan 

Adjust the calibration curve if necessary based on the internal standard 

Calculate the peak areas based upon assumed peak start and peak end parameters, and calculate peak center masses through another layer of peak centroid calculation (peak centroiding) with assumed peak definition parameters.

This approach suffers from noise and baseline interferences for low signals, asymmetrical peak shapes for high signals, the changes in peak shapes from time to time, isotope interferences from unit mass resolution systems, and the intrinsic errors in peak centroiding through the use of only part of the mass spectral peak data in the calculation. 

Figure 1. A simplified flow diagram for HAMSCA.			

A new approach (referred to here as the highly accurate mass spectral calibration approach, or HAMSCA) addresses all these and other factors in mass spectral data analysis such that all systematic variations are calibrated out through the use of sufficient external and internal standards. No peak distortions or other artifacts are introduced by using a comprehensive mathematical data analysis process to achieve optimal signal processing and averaging (maximizing 

 in the above equation), and the quantitative accuracy (peak area integration) and qualitative accuracy (mass accuracy) are limited only by random errors in ion detection. This new method can be summarized as follows: 					

Statistically examine peak shapes and mass errors for the calibration standard

Build a correction to compensate for any mass errors and peak asymmetries 

Apply the correction to compensate for both mass errors and peak shape asymmetries

Update the correction if necessary based upon the internal standard

Find peak areas and peak center masses using a parameter-free peak picker without the use of any assumptions.

Test runs to validate HAMSCA were performed on a variety of instrument types from different vendors. Both low resolution (0.5-5 Da FWHM) and higher resolution instruments (~0.1 Da FWHM) were tested. First an instrument calibration standard is run on the instrument. The data is transferred to the computer program, which processes the data using HAMSCA. Key peaks are designated, and the software calculates the correction function for the spectrometer automatically. Next the sample is run in the normal fashion using the normal instrument vendor-provided software. An internal standard is included when seeking maximum mass accuracy. The sample spectrum then is transferred back to the program, which processes the raw spectrum to obtain the fully calibrated and aligned spectrum. Finally, a mathematically sound peak analysis algorithm is applied to the processed spectrum to yield high mass accuracy and unbiased peak picking results. 			

Figure 2 depicts the results obtained from a typical lower resolution instrument. An external standard of sodium trifluoroacetate (STFA) solution was used to develop the correction function. A sample of terfenadine was run along with internal standards of reserpine and promethazine. The monoisotopic mass was calculated to be 472.3191 versus the actual mass of 472.3216. This is about 5 ppm in mass accuracy as compared with the typical observed error of ~500 ppm in normal instrument operation. In addition, it should be noted that there is substantial enhancement in S/N (about 3X) as well as improvements in peak shape and symmetry which greatly enhances the ability to perform automated peak detection.

Figure 2. A mass accuracy of 5 ppm is achieved on Sciex Q Trap 4000 instrument of unit mass resolution for the drug compound terfenadine (exact monoisotopic mass 472.3216). 			

Figure 3 depicts the results from a higher resolution instrument with a typical observed mass accuracy of 5 ppm (four-scan average) using the vendor-provided software routines. In this case, both an internal and external standard was used in the calibration. The mass accuracy was improved in this case to 2.5 ppm when using HAMSCA. Again note the substantial improvement in peak shape and noise as well.

Figure 3. From top to bottom shows processing before and after HAMSCA. Sample: polyalanine solution; internal standard: 8-alanine ion (C24H43N8O9+, monoisotopic mass 587.3153); external standard: polyalanine standard; instrument: Waters Micromass qTOF II. 			

Figure 4 is an example of a low-resolution system run using only an external standard. While the best mass accuracy is obtained using internal standard, in this case it was not of major concern. It can be seen that the HAMSCA method of processing still offers substantial improvements in S/N for improved detection limits and peak shape for improved quantitative accuracy.   

Figure 4. In addition to mass accuracy improvement, significant improvement in S/N is gained as well (3X or greater). Sample: sodium trifluoroacetate (STFA) solution; internal standard: none; external standard: STFA standard; instrument: Sciex Q Trap 4000.			

To validate the operability of HAMSCA across different instrument designs and vendors, the authors have attempted to measure results on as large a cross section of available instruments as possible. Table I illustrates a subset of some of the results obtained. 

It has been shown that current MS processing techniques typically fall far short of extracting the maximum amount of information from the data. Significant improvements in MS S/N and peak shape are obtained by simply running an external standard and using it as a reference to calculate a mathematically accurate correction function. With the addition of an internal standard, HAMSCA can routinely extend the capabilities of unit mass resolution instruments to obtain mass accuracies of up to 5 ppm. This is approximately 100X better than results typically obtained and suitable for accurate compound identification, a capability usually available only on more expensive systems. This might make it possible for relatively low-cost instruments to be used routinely for biomarker discovery, metabonomics, and proteomics research. 			

Substantial improvements also are seen on higher resolution spectrometers in S/N, peak shape, and mass accuracy. Although mass accuracy improvements are not as dramatic on higher end instruments, the improvements still are substantial and can improve the results in the above mentioned application areas. However, since these instruments typically are purchased to perform high mass accuracy measurements, the resulting improvements might be of greater value for users of these systems by further increasing the accuracy and confidence in the data.  			

Finally, due to the fact that the algorithms used are linear operations, the speed of processing could allow the results to be calculated in real-time on vendor supplied software.  

 are with Cerno Bioscience (New Haven, CT; 

The exploding field of proteomics has highlighted the need to improve the performance of mass spectrometry, both quantitatively and qualitatively. These needs have led instrument manufacturers to produce instruments of increasingly higher quality, but little work has been done to address the fundamental errors inherent in the measurement technique itself. This paper examines these errors and demonstrates that the appropriate mathematical correction of MS data can significantly improve the performance of both low- and high-resolution instruments.

Ming Gu

Articles

Enhancing Mass Spectral Formula Determination by Heuristic Rules

Calibration Stability for Formula Determination on a Single-Quadrupole GC–MS System

A Mathematical Approach to Error Reduction in Mass Spectrometry