Forensic Identiﬁcation of Gender from Fingerprints
Crystal Huynh, Erica Brunelle, Lenka Halámková, Juliana Agudelo, and Jan Halámek*
Department of Chemistry, University at Albany, State University of New York, 1400 Washington Avenue, Albany, New York 12222,
ABSTRACT: In the past century, forensic investigators have universally accepted
ﬁngerprinting as a reliable identiﬁcation method, which relies mainly on pictorial
comparisons. Despite developments to software systems in order to increase the probability
and speed of identiﬁcation, there has been limited success in the eﬀorts that have been made
to move away from the discipline’s absolute dependence on the existence of a prerecorded
matching ﬁngerprint. Here, we have revealed that an information-rich latent ﬁngerprint has
not been used to its full potential. In our approach, the content present in the sweat left
behindnamely the amino acidscan be used to determine physical such as gender of the
originator. As a result, we were able to focus on the biochemical content in the ﬁngerprint using a biocatalytic assay, coupled with
a specially designed extraction protocol, for determining gender rather than focusing solely on the physical image.
In modern criminology, ﬁngerprints play a major role in
forensics as a means of identiﬁcation.
In the past 110 years,
the development of ﬁngerprint analysis has stalled at simple
visual comparison and matching, even though ﬁngerprintsas
samples of biological origin analogous to other body ﬂuids
have the potential to provide much more information.
Currently, the shape, size, and unique patterns associated
with ﬁngerprints are compared using various computational
programs, such as Automated Fingerprint Identiﬁcation System
(AFIS). However, the ultimate setback is the requirement of
matching ﬁngerprints to be stored in a database or for the
person of interest to be physically present for comparison. If
neither of these conditions is fulﬁlled, the print is reduced to
merely exclusionary evidence,
despite being stored in a
separate database for future use when comparing them with
new, incoming ﬁngerprints. The same can be said about DNA
analysis. Even though DNA can provide potentially the most
signiﬁcant information about the ﬁngerprint originator, DNA
analysis can take weeks or months to be processed, and if there
is no matching DNA proﬁle in the database, then the potential
for information is signiﬁcantly reduced. The purpose of our
approach is to address the issue of a ﬁngerprint not having an
immediate matching image or DNA proﬁle. While the patterns
of individual ﬁngerprints, as well as DNA, act as important
sources of evidence, it is often overlooked that sweat and
sebum are also left behind with a ﬁngerprint. Sweat contains
varying amounts of metabolites produced by the body
depending on various processes related to metabolism.
Metabolism, which is directed by a combination of multiple
hormone-based control mechanisms,
acts as a function of
physical properties such as gender, age, ethnicity, or health
status. It is also known that the amino acid content of an
individual can slightly vary depending on the physiological state
of the individual’s metabolism. This slight variation occurs over
time scales of several hours and can also be somewhat aﬀected
by certain medications
as well as after the consumption of
certain foods. It has been found that amino acid levels diﬀer
between people in diﬀerent demographic groups such as gender
(i.e., male and female).
As a result of this diﬀerence, the
biological/biochemical content from the sweat left behind with
ﬁngerprints can be exploited, in a method similar to the way
clinical diagnostics uses other body ﬂuid contents,
valuable information on diﬀerent “persons of interest”
connected to a particular crime scene.
In fact, research has already started in this area using
spectrometric methods such as matrix-assisted laser desorption
ionization (MALDI) and liquid chromatography−mass spec-
trometry (LC−MS) as well as spectroscopic methods including
infrared and Raman spectroscopy.
Two examples of this
research involve the use of desorption electrospray ionization
(DESI) mass spectrometry to detect various explosive-related
and Raman spectroscopy to detect drug
compounds in the secretions left behind with ﬁngerprints.
At the moment these methods require rather sophisticated
instrumentation that do not always allow for on-site analysis,
despite their incredibly selective and reliable nature. These
devices also demand highly specialized personnel that are not
likely to be available as part of the immediate forensic
investigation team. In addition, recently published eﬀorts
using nanostructured materials and aﬃnity-based techniques
aim for the detection of compounds such as cotinine
in the latent ﬁngerprint content. Notably, there has
been progress in the use of various instrumental studies on
latent ﬁngerprints as demonstrated by the Kazarian group.
They have performed studies in spectroscopic
of latent ﬁngerprints. The possibility of using the
endogenous compounds in ﬁngerprints for the determination
of various personal attributes has also been noted by other
The vast majority of these methods,
however, can only focus on the presence or absence of certain
Received: August 30, 2015
Accepted: October 13, 2015
© XXXX American Chemical Society ADOI: 10.1021/acs.analchem.5b03323
Anal. Chem. XXXX, XXX, XXX−XXX
chemical compounds. While this can be useful, there is still a
limit to the conclusions that can be drawn based on these
Despite the success in diﬀerentiating between genders using
ridge density, a complex statistical analysis as well as visual
comparison by counting the number of ridges present in a 25
mm2area that are present is still required.
We have recently
demonstrated that bioaﬃnity-based analytical systems are able
to distinguish between African-American and Caucasian
ethnicities by analysis of biomarkers commonly present in
This was followed by a similar study that showed it is
also possible to distinguish between males and females.
very recent example
of the success of these bioaﬃnity systems
involves the parallel analysis of two blood markers in order to
determine age of the blood sample found at a crime scene.
Because the contents of ﬁngerprints can provide information in
a way that is analogous to the contents of body ﬂuids, it should
also be able to provide information similar to the previously
mentioned study. In this work, the proposed methodology
utilizes the bioaﬃnity interaction between an enzyme and its
ligand (for example, substrate, cosubstrate, activator, inhibitor,
etc.) in order to generate a visible color change that can be seen
by the naked eye or spectroscopically quantiﬁed. The
extraction/bioassay system presented here is the ﬁrst proof-
of-concept example of a system that can detect a physical
characteristic of the ﬁngerprint originator (male/female origin)
based on the chemical content of the ﬁngerprint, ignoring the
traditional pictorial approach.
Ethics Statement. The Institutional Review Board, Oﬃce
of Pre-Award and Compliance at the University at Albany has
fully approved the experimental protocols demonstrated in this
manuscript. These protocols were carried out in accordance to
the oﬃce’s requirement of obtaining informed consent, in the
form of a signature from each volunteer, acknowledging that
they are aware of the procedure that will take place, any risks or
beneﬁts that may accompany the study, as well as acknowl-
edging that they will not receive any payment for their
participation. Informed consent from all volunteers who
participated in this research study was obtained.
Enzymatic Assay Components. The following enzymes
and organic/inorganic chemicals were purchased from Sigma-
Aldrich: L-amino acid oxidase type IV (L-AAO, E.C. 188.8.131.52),
horseradish peroxidase type VI (HRP, 184.108.40.206), o-dianisidine,
(+)-sodium L-ascorbate, triethanolamine (TEA), L-aspartic acid,
L-threonine, L-serine, L-glutamic acid, L-asparagine, L-glutamine,
L-cysteine, L-proline, glycine, L-alanine, L-valine, L-cystine, L-
methionine, L-isoleucine, L-leucine, L-tyrosine, L-phenylalanine,
β-alanine, L-ornithine, L-lysine, L-tryptophan, L-histidine, L-
arginine, and L-citrulline. Water used in all of the experiments
was ultrapure (18.2 MΩ·cm) water from PURELAB ﬂex, an
ELGA water puriﬁcation system.
Detection of Amino Acids. The dual-enzyme cascade,
displayed in Scheme 1, was designed and optimized in the
present study and realized in pH 7.6 TEA buﬀer containing 20
mU of L-AAO, 3 U of HRP, and 85 μMo-dianisidine. The
cascade is activated when L-AAO reacts with a range of
concentrations of amino acids present in the sample, which
results in the conversion of O2to H2O2. The HRP then
consumes the H2O2, causing the oxidation of the dye, o-
dianisidine, present in the system. This results in the formation
of the oxidized form of o-dianisidine, which is observed
spectroscopically at λ= 436 nm. The intensity of visible color
production is then proportional to the amino acid concen-
trations present in the sample. This enzyme cascade was ﬁrst
optimized using mixtures of the average amino acid
concentrations speciﬁc for males and females, respectively.
The reactions and optical measurements were performed at 37
°C using a SpectraMax Plus384 (Molecular Devices, CA)
microplate reader with polystyrene (96 well) microtiter plates.
The signal corresponding to the concentration of oxidized o-
dianisidine was measured optically at λ= 436 nm.
Statistical Analysis. R-project software
was used to
generate randomized concentrations of each of the amino acids
found in ﬁngerprint content to create 50 amino acid mixtures,
with 25 mixtures representing male samples and 25 mixtures
representing female samples. The L-AAO/HRP biocatalytic
assay was then performed using the 25 mimicked male and 25
mimicked female ﬁngerprint samples. Receiver operating
characteristic (ROC) analysis was used to evaluate the
performance of the assay and estimate the probability of
distinguishing between the male (25 samples) and female (25
samples) groups of human ﬁngerprint content. Using ROC
analysis, the threshold (above which the absorbance changes
correspond to the female group) that yielded the maximum
accuracy was determined. ROC analysis involves changing the
threshold and observing the eﬀect on the predictive power of
the model to produce an ROC curve. The area under the ROC
curve (AUC) is a single measurement that summarizes the
overall discriminating ability of the assay. It represents the
probability that the diagnostic test will correctly distinguish
between the male and female samples. The larger the AUC, the
higher the probability that each sample will be identiﬁed
correctly. Lastly, to demonstrate the viability of this method,
real ﬁngerprints were analyzed using the bioassay.
Amino Acid Extraction from Authentic Fingerprints.
For the purpose of determining the viability of the biocatalytic
assay when used with authentic samples, real ﬁngerprint
samples were collected from volunteers and the amino acids
were extracted with a new protocol that was developed during
this project. This protocol combines the use of an elevated
temperature and acidic conditions to extract the water-soluble
amino acids from the lipid-based content of the ﬁngerprint,
which is composed of triglycerides, wax esters, free fatty acids,
The aqueous phase containing the amino acids
was removed from the ﬁngerprint surface and subsequently
analyzed by the proposed biocatalytic assay. The extraction
process consisted of the following steps: the ﬁngerprints were
deposited onto a portable surface composed of polyethylene
and 120 μL of 0.01 M HCl was placed directly onto the
ﬁngerprint, covering an area of 63 mm2. The entire surface was
Scheme 1. Dual-Enzyme Cascade Assay Containing L-AAO
and HRP Used for the Diﬀerentiation of Gender via
The abbreviations used are L-AAO (L-amino acid oxidase) and HRP
Analytical Chemistry Article
Anal. Chem. XXXX, XXX, XXX−XXX
then heated at 40 °C for 20 min. This process causes the amino
acid content in the ﬁngerprint to migrate from the lipid-based
content into the aqueous acidic solution, while the lipid-based
content remains on the portable lipophilic surface. The aqueous
acidic solution was then collected oﬀof the portable surface
and used as the sample for analysis.
Gender Determination via Bioassay. Here, we are
proposing a bioaﬃnity-based sensing system that has the
capability to distinguish between ﬁngerprint samples from
males and females using the concentrations of amino acids
present in ﬁngerprint content. The logic for this determination
is based on the proven fact that females have diﬀerent
concentrations of amino acids in their systems than
The L-AAO enzyme has the ability to use a
large range of L-amino acids as substrates, with a varying degree
which drives the enzyme to convert oxygen to
peroxide. L-AAO has previously been used as an amperometric
for the identiﬁcation of amino acids as well as in an
assay for detecting intestinal peptide hydrolase activity.
peroxide then acts as a substrate for the secondary element of
the cascade, HRP, which oxidizes a dye acting as a cosubstrate,
thus, producing a signal at a particular wavelength. The dye
used in this case was o-dianisidine, which becomes oxidized by
HRP and is consequently detected at 436 nm.
below, this method, combined with the newly developed
extraction protocol allowing for the isolation of water-soluble
amino acids from the lipid-based ﬁngerprint content, requires
only a minuscule amount of substrate. More importantly, it
provides a quick male/female response and can be performed
on-site. These results can narrow down the possibilities in a
suspect pool in a quick and timely manner when there is no
matching ﬁngerprint image or DNA proﬁle in the correspond-
ing databases. Furthermore, this type of analysis can potentially
be utilized by any and all members of law enforcement with no
need for specialized training, as it works in a similar manner to
pregnancy strips or glucometers.
Statistical Analysis of Mimicked Samples. The dis-
tribution of amino acid concentrations in human ﬁngerprint
content was previously studied,
and a list of the
corresponding average amino acid concentrations can be
found in Table 1. In order to investigate the distribution of
and variability in amino acid concentrations found in
ﬁngerprint content, the inﬂuence of gender was examined.
The studies, referenced above, reported signiﬁcant diﬀerences
in the concentrations of certain amino acids when comparing
genders. The reported data were used in the present study to
prepare solutions mimicking the levels of all amino acids
present in the ﬁngerprints of diﬀerent genders.
The ﬁrst step of our study began with the statistical analysis
of the available data from previous studies. The values were not
normally distributed, but rather positively skewed and
consistent with a log-normal distribution. The parameters of
the log-normal distribution were available only for overall
amino acid concentrations, while the distribution parameters
estimated from the male and female data came from
logarithmic untransformed data. The available parameters for
a normal distribution were ﬁrst corrected for a log-normal
distribution. For each of the 23 amino acids present in the
ﬁngerprint content, random values according to the recalcu-
lated parameters of the log-normal distribution in males and
females were generated using R-project software.
sets of concentration values (25 for males and 25 for females)
of all 23 amino acids were produced and randomly grouped
together. This allowed for the concentration groups to capture
a range of concentrations that amino acids can take in males
and females. Then 25 experiments using these concentration
groups were performed for each gender.
After obtaining the optical responses for the model solutions,
seen in Figure 1, the measured output signal was deﬁned as the
absorbance of the oxidized dye as a function of time, once the
reaction was initiated. The optical responses were then used for
further statistical analysis. The bottom part of the graph
corresponds to the male samples (lower concentrations of the
amino acids) while the top part of the graph agrees with female
samples (higher concentrations of the amino acids). The
Table 1. Average Amino Acid Concentration (mM) Values
for Males and Females Derived from Sweat
AA female conc. (mM) male conc. (mM)
Thr 0.2090 0.1121
Ser 0.9840 0.5208
Glu 0.1780 0.1109
Gly 0.6463 0.3418
Ala 0.3870 0.1968
Cit 0.1967 0.1267
Asp 0.1196 0.0638
Asn 0.0380 0.0161
Gln 0.0178 0.0120
Pro 0.0728 0.0349
Val 0.0919 0.0459
Cys 0.0012 0.0009
Met 0.0085 0.0034
Iso 0.0494 0.0229
Leu 0.0625 0.0324
Tyr 0.0559 0.0303
Phe 0.0378 0.0172
β-ala 0.0128 0.0034
Orn 0.1361 0.0684
Lys 0.0528 0.0285
Trp 0.0151 0.0071
His 0.1790 0.0804
Arg 0.0948 0.0540
These values have been previously reported and were used to prepare
mimicked ﬁngerprint samples.
Figure 1. Change in signal response (λ= 436 nm) corresponding to
the oxidation of o-dianisidine upon reaction of the analytical system.
The top red traces indicate the female samples, and the bottom blue
traces indicate the male samples.
Analytical Chemistry Article
Anal. Chem. XXXX, XXX, XXX−XXX
produced responses, which relate the absorbance to the
concentrations of all 23 amino acids, are clearly diﬀerent
between males and females, with a small overlap between them.
The readout time was initially set to 420 s; however, after
repeating the statistical analysis using later readout times up to
1800 s, the ﬁnal result remained almost unchanged. The amino
acid concentrations were chosen to follow the published
distributions that are relevant for males and females; thus, real
applications should generate signals that correspond to the
signal distribution of the mimicked samples.
The area under the ROC curve,
also known as AUC, was
estimated by the trapezoidal integration method, and the
corresponding 95% conﬁdence interval (CI) was estimated
using the method described by DeLong et al.
The AUC was
estimated at 0.99 (95% CI, 0.98−1.00) from the ROC curve,
Figure 2, which means that the L-AAO/HRP assay has a 99%
probability of correctly diﬀerentiating between male and female
ﬁngerprints. The ROC curve was generated from absorbance
changes, and the best absorbance threshold of 0.439, which
balanced the trade-oﬀthat exists between sensitivity and
speciﬁcity, was determined. The absorbance change is the most
accurate cutoﬀpoint for discrimination between male and
female ﬁngerprint samples. As shown, ROC analysis has proven
the potential of this assay to correctly diﬀerentiate between
samples from males and females.
Evaluation of Authentic Fingerprint Samples. The
success in distinguishing between genders utilizing the bioassay
explained above on the mimicked samples led to further studies
involving authentic latent ﬁngerprints. These ﬁngerprints were
collected from two groups of volunteers, Caucasian males and
Caucasian females. The ﬁngerprints were collected on poly-
ethylene ﬁlm according to an established procedure described
by Croxton et al.
Once the ﬁngerprints were deposited, the
extraction protocol, detailed in the Experimental Section, was
applied. Following the extraction, the amino acid samples were
subjected to the same bioassay as the mimicked samples. As
anticipated, the aqueous solution removed from the ﬁngerprints
contained the amino acids necessary for analysis, while the
nonpolar content remained on the polyethylene ﬁlm. The
success of the extraction of amino acids as well as the bioassay’s
performance is demonstrated in Figure 3. As seen here, the
samples from the authentic ﬁngerprints generated a noticeably
lower optical signal than that of the mimicked samples (Figure
1), which is a result of the dilution that occurs during the
extraction of the amino acids from the authentic ﬁngerprints.
However, despite the diﬀerence in signal intensity, there is still
a signiﬁcant diﬀerence between the samples obtained from
males and the samples obtained from females. The results of
the L-AAO/HRP assay using authentic ﬁngerprints were
consistent with the results of the L-AAO/HRP assay using
mimicked samples in that the female samples generated a
signiﬁcantly higher optical signal than that of the male samples,
corresponding to the higher amino acid content. Additionally,
the relative standard deviation between the signals produced by
the authentic ﬁngerprint samples, both left and right thumbs,
does not exceed 8%.
Evaluation of Various Extraction Surfaces. Following
the identiﬁcation of gender from the authentic ﬁngerprint
samples, the extraction protocol and biocatalytic assay were
further tested on various surfaces that could be found at a crime
scene. Given that authentic male ﬁngerprints did not generate a
signiﬁcant signal, only female ﬁngerprints were used for this
experiment. Three female ﬁngerprints were deposited onto
multiple surfaces including a door knob, a laminate desktop, a
composite benchtop, and a computer screen. The polyethylene
ﬁlm used in the previous experiment for the determination of
gender was used to remove the ﬁngerprints from the respective
surfaces. The samples were then subjected to the same
extraction protocol, described above, as well as the L-AAO/
Figure 2. Trade-oﬀbetween sensitivity and speciﬁcity is shown by
presenting data as a receiver operating characteristic (ROC) curve.
Area under the ROC curve (AUC) is 99%, which is the probability for
the presented assay to correctly distinguish between males and females
based on the amino acids’concentrations in the respective ﬁngerprint
samples. The optimum cutoﬀpoint was chosen with a sensitivity and
speciﬁcity of 96%. Random choice is denoted by the gray diagonal line.
Figure 3. Data obtained from authentic ﬁngerprint samples from Caucasian males and females: (a) the left thumb and (b) the right thumb.
Analytical Chemistry Article
Anal. Chem. XXXX, XXX, XXX−XXX
HRP bioassay. Figure 4 demonstrates that the extraction
protocol and the bioassay are capable of identifying a female
ﬁngerprint, regardless of the surface from which it was taken.
The L-AAO/HRP bioassay described above for the purpose of
distinguishing between ﬁngerprint samples obtained from both
males and females has proven to be reliable and reproducible.
The ROC analysis conducted using 50 mimicked ﬁngerprint
samples generated statistics proving that it is possible to
determine the gender of the ﬁngerprint originator using this
method. Initial experiments using mimicked ﬁngerprint
methods, concluded that there was a 99% chance of
determining the correct gender of the ﬁngerprint originator.
Furthermore, a reliable sample extraction protocol was
employed for the extraction of the necessary substrates
amino acidsfrom real ﬁngerprint samples collected from
Caucasian male and female volunteers. In the case mentioned
above, the thumbprints of the volunteers were deposited onto a
portable polyethylene ﬁlm, where they were subjected to acidic
conditions. The amino were then separated from the lipid-
based components through heating the polyethylene ﬁlm. The
results from the analysis of authentic ﬁngerprint samples further
demonstrated the ability of the bioassay to diﬀerentiate
between male and female ﬁngerprint samples based on the
signiﬁcant diﬀerence in absorbance intensities. In addition, the
durability of the bioassay and extraction process was
successfully determined using various surfaces from which the
ﬁngerprints were collected.
The authors declare no competing ﬁnancial interest.
(1) Yamashita, B.; French, M. Fingerprint Sourcebook; National
Institute of Justice/NCJRS: Rockville, MD, 2010.
(2) Lennard, C. Aust. J. Forensic Sci. 2013,45, 356−367.
(3) Francese, S.; Bradshaw, R.; Ferguson, L. S.; Wolstenholme, R.;
Clench, M. R.; Bleay, S. Analyst 2013,138, 4215−4228.
(4) Thody, A. J.; Shuster, S. Physiol. Rev. 1989,69, 383−416.
(5) Kutyshenko, V. P.; Molchanov, M.; Beskaravayny, P.; Uversky, V.
N.; Timchenko, M. A. PLoS One 2011,6, e28824.
(6) Hier, S. W.; Cornbleet, T.; Bergeim, O. J. Biol. Chem. 1946,166,
(7) Coltman, C. A., Jr.; Rowe, N. J.; Atwell, R. J. Am. J. Clin. Nutr.
(8) Croxton, R. S.; Baron, M. G.; Butler, D.; Kent, T.; Sears, V. G.
Forensic Sci. Int. 2010,199,93−102.
(9) Khan, M. S.; Thouas, G.; Shen, W.; Whyte, G.; Garnier, G. Anal.
Chem. 2010,82, 4158−4164.
(10) Corstjens, A. M.; Ligtenberg, J. J. M.; van der Horst, I. C. C.;
Spanjersberg, R.; Lind, J. S. W.; Tulleken, J. E.; Meertens, J. H. J. M.;
Zijlstra, J. G. Crit. Care 2006,10, R135.
(11) Haese, A.; de la Taille, A.; van Poppel, H.; Marberger, M.;
Stenzl, A.; Mulders, P. F. A.; Huland, H.; Abbou, C. m.-C.; Remzi, M.;
Tinzl, M.; Feyerabend, S.; Stillebroer, A. B.; van Gils, M. P. M. Q.;
Schalken, J. A. Eur. Urol. 2008,54, 1081−1088.
(12) Streckfus, C. F.; Bigler, L. R. Oral Dis. 2002,8,69−76.
(13) Hazarika, P.; Russell, A. D. Angew. Chem., Int. Ed. 2012,51,
(14) Ifa, D. R.; Manicke, N. E.; Dill, A. L.; Cooks, R. G. Science 2008,
(15) Day, J. S.; Edwards, H. G. M.; Dobrowski, S. A.; Voice, A. M.
Spectrochim. Acta, Part A 2004,60, 1725−7130.
(16) Leggett, R.; Lee-Smith, E. E.; Jickells, S. M.; Russell, D. A.
Angew. Chem., Int. Ed. 2007,46, 4100−4103.
(17) Hazarika, P.; Jickells, M. S.; Wolff, K.; Russell, A. D. Angew.
Chem., Int. Ed. 2008,47, 10167−10170.
(18) Ricci, C.; Bleay, S.; Kazarian, S. G. Anal. Chem. 2007,79, 5771−
(19) Ricci, C.; Phiriyavityopas, P.; Curum, N.; Chan, K. L.; Jickells,
S.; Kazarian, S. G. Appl. Spectrosc. 2007,61, 514−522.
(20) Bailey, M. J.; Bright, N. J.; Croxton, R. S.; Francese, S.;
Ferguson, L. S.; Hinder, S.; Jickells, S.; Jones, B. J.; Jones, B. N.;
Kazarian, S. G.; Ojeda, J. J.; Webb, R. P.; Wolstenholme, R.; Bleay, S.
Anal. Chem. 2012,84, 8514−8523.
(21) Acree, M. A. Forensic Sci. Int. 1999,102,35−44.
(22) Gungadin, S. Internet J. Med. Update 2007,2,1−4.
(23) Kramer, F.; Halámková, L.; Poghossian, A.; Schöning, M. J.;
Katz, E.; Halámek, J. Analyst 2013,138, 6251−6257.
(24) Bakshi, S.; Halámková, L.; Halámek, J.; Katz, E. Analyst 2014,
(25) Agudelo, J.; Huynh, C.; Halámek, J. Analyst 2015,140, 1411−
(26) Michalski, S.; Shaler, R.; Dorman, F. L. J. Forensic Sci. 2013,58,
(27) Mark, H.; Harding, C. R. Int. J. Cosmet. Sci. 2013,35, 163−168.
(28) Hadorn, B.; Hanimann, F.; Anders, P.; Curtius, H. C.;
Halverson, R. Nature 1967,215, 416−417.
(29) Liappis, N.; Jäkel, A. Arch. Dermatol. Res. 1975,254, 185−203.
(30) Pierce, C. M. The Detailed Composition of Feces, Sweat, and
Urine: An Annotated Bibliography; Lockheed Missiles and Space
Company: Sunnyvale, CA, 1962.
(31) Hamilton, P. B. Nature 1965,205, 284−285.
(32) Geueke, B.; Hummel, W. Enzyme Microb. Technol. 2002,31,
(33) Du, X.-Y.; Clemetson, K. J. Toxicon 2002,40, 659−665.
(34) Wellner, D.; Meister, A. J. Biol. Chem. 1960,235, 2013−2018.
(35) Kwan, R. C. H.; Chan, C.; Renneberg, R. Biotechnol. Lett. 2002,
(36) Nicholson, J. A. K.; Kim, Y. S. Anal. Biochem. 1975,63, 110−
Figure 4. Female ﬁngerprints (n= 3) taken from ﬁve diﬀerent
surfaces. Inset bar diagram represents the change in absorbance at λ=
436 nm, after 600 s of assay completion. Surfaces A−E in the bar
diagram represent a computer screen, laminate desktop, door knob,
polyethylene ﬁlm, and a composite benchtop, respectively.
Analytical Chemistry Article
Anal. Chem. XXXX, XXX, XXX−XXX
(37) Claiborne, A.; Fridovich, I. Biochemistry 1979,18, 2324−2329.
(38) de Puit, M.; Ismail, M.; Xu, X. J. Forensic Sci. 2014,59, 364−370.
(39) R Development Core Team R: A Language and Environment for
Statistical Computing; R Foundation for Statistical Computing: Vienna,
(40) Gentleman, R.; Ihaka, R. R-project.https://www.r-project.org
(41) Zweig, M. H.; Campbell, G. Clin. Chem. 1993,39, 561−577.
(42) DeLong, E. R.; DeLong, D. M.; Clarke-Pearson, D. L. Biometrics
Analytical Chemistry Article
Anal. Chem. XXXX, XXX, XXX−XXX