39
Nat Methods. 2012 Jul 15;9(8):796-804.
Wisdom of crowds for robust gene network
inference.
Marbach D, Costello JC, Küfner R, Vega NM, Prill RJ, Camacho DM, Allison KR; DREAM5
Consortium, Kellis M, Collins JJ, Stolovitzky G. Aderhold A, Allison KR, Bonneau R, Camacho
DM, Chen Y, Collins JJ, Cordero F, Costello JC, Crane M, Dondelinger F, Drton M, Esposito R,
Foygel R, de la Fuente A, Gertheiss J, Geurts P, Greenfeld A, Grzegorczyk M, Haury AC, Holmes
B, Hothorn T, Husmeier D, Huynh-Thu VA, Irrthum A, Kellis M, Karlebach G, Küfner R, Lèbre
S, De Leo V, Madar A, Mani S, Marbach D, Mordelet F, Ostrer H, Ouyang Z, Pandya R, Petri T,
Pinna A, Poultney CS, Prill RJ, Rezny S, Ruskin HJ, Saeys Y, Shamir R, Sîrbu A, Song M, Soranzo N,
Statnikov A, Stolovitzky G, Vega N, Vera-Licona P, Vert JP, Visconti A, Wang H, Wehenkel L,
Windhager L, Zhang Y, Zimmer R.
Abstract
Reconstructing gene regulatory networks from high-throughput data is a long-standing
challenge. Through the Dialogue on Reverse Engineering Assessment and Methods (DREAM)
project, we performed a comprehensive blind assessment of over 30 network inference
methods on
Escherichia coli, Staphylococcus aureus, Saccharomyces cerevisiae
and
in silico
microarray data. We characterize the performance, data requirements and inherent biases
of diferent inference approaches, and we provide guidelines for algorithm application and
development. We observed that no single inference method performs optimally across all
data sets. In contrast, integration of predictions from multiple inference methods shows
robust and high performance across diverse data sets. We thereby constructed high-conf-
dence networks for E. coli and S. aureus, each comprising ~1,700 transcriptional interactions
at a precision of ~50%. We experimentally tested 53 previously unobserved regulatory inte-
ractions in
E. coli
, of which 23 (43%) were supported. Our results establish community-based
methods as a powerful and robust tool for the inference of transcriptional gene regulatory
networks.
Forensic Sci Int. 2012 Jun 10;219(1-3):64-75
.
Raman spectroscopy and laser desorption mass
spectrometry for minimal destructive forensic
analysis of black and color inkjet printed documents.
Heudt L, Debois D, Zimmerman TA, Köhler L, Bano F, Partouche F, Duwez AS, Gilbert B,
De Pauw E.
Abstract
Inkjet ink analysis is the best way to discriminate between printed documents, or even though
more difcult, to connect an inkjet printed document with a brand or model of printers.
Raman spectroscopy and laser desorption mass spectrometry (LDMS) have been demons-
trated as powerful tools for dyes and pigments analysis, which are ink components. The aim
of this work is to evaluate the aforementioned techniques for inkjet inks analysis in terms of
discriminating power, information quality, and nondestructive capability. So, we investigated
10 diferent inkjet ink cartridges (primary colors and black), 7 from the HP manufacturer and
one each from Epson, Canon and Lexmark. This paper demonstrates the capabilities of three
methods: Raman spectroscopy, LDMS and MALDI-MS. Raman spectroscopy, as it is preferable
to try the nondestructive approach frst, is successfully adapted to the analysis of color prin-
ted documents in most cases. For analysis of color inkjet inks by LDMS, we show that a MALDI
matrix (9-aminoacridine, 9AA) is needed to desorb and to ionize dyes from most inkjet inks
(except Epson inks). Therefore, a method was developed to apply the 9AA MALDI matrix
directly onto the piece of paper while avoiding analyte spreading. The obtained mass spectra
are very discriminating and lead to information about ink additives and paper compositions.
Discrimination of black inkjet printed documents is more difcult because of the common
use of carbon black as the principal pigment. We show for the frst time the possibility to
discriminate between two black-printed documents coming from diferent, as well as from
the same, manufacturers. Mass spectra recorded from black inks in positive ion mode LDMS
detect polyethylene glycol polymers which have characteristic mass distributions and end
groups. Moreover, software has been developed for rapid and objective comparison of the
low mass range of these positive mode LDMS spectra which have characteristic unknown
peaks.