Biology Reference
In-Depth Information
42%. They conclude that the major problem lies
with divergent assignment of organism and
splice variants and differential methods of
dealing with multiprotein complexes.
Both reports highlight the need for a standard-
ized format for submitting protein
e
protein
interaction data to the databases. There is now
an effort through the IMEx consortium aiming
to systematically address this problem by setting
guidelines for submitting and curating protein
e
protein interaction data.
53,54
With the many different databases harboring
protein
e
protein interaction data, it can be diffi-
number of both false positive and negative
entries is signi
cant. One of the main factors
contributing to this noise in the databases is the
experimental methodologies used for identifying
interacting proteins. The two most commonly
utilized high-throughput methods are: (1) yeast
2-hybrid and (2) af
nity puri
cation followed
by tandem MS (MS
2
).
Yeast 2-Hybrid Method
Classical yeast 2-hybrid studies rely on the
fusion of protein fragments to components of
the yeast transcriptional machinery so that an
interaction between two protein fragments will
bring different transcriptional proteins into suffi-
-
cult to go through all of them to explore the
available data for an individual protein, let alone
a list of proteins. There are several publicly avail-
able web-based tools such as STRING,
55
UniHI,
56
and iRefWeb
57
that allow for simulta-
neous query of protein
e
protein interaction
data from multiple databases at the same time.
In addition to the publicly available tools, there
exists proprietary software such as Ingenuity
Pathway Analysis (IPA)
-
cient proximity to initiate the expression of
a reporter gene. This method was developed in
1989 by Fields and Song
58
and has been used
extensively. However, as the authors recognized
from the beginning, this system requires the
interaction to take place in the nucleus so that
the fusion proteins do not impede the function
of the transcriptional machinery through steric
interference and the proteins of interest do not
individually activate transcription. These
requirements mean that the study of transcrip-
tion factors, membrane proteins, and proteins
that are localized in other regions of the cell
can be highly problematic.
Since the initial development of the yeast
2-hybrid method, several permutations have
been developed that all rely on the interactions
between a bait and prey protein to bring together
two complementary parts of some portion of the
cell machinery that can be linked to a measurable
output. Many of these methods were reviewed
by Bruckner et al.
59
Bruckner noted that the
selection of the appropriate yeast 2-hybrid
method based on the type of proteins to be
studied is essential (i.e., nuclear, cytoplasmic,
membrane, etc.). However, this selection does
not ensure that many interactions that occur
naturally will not be detected due to steric
hindrance that
from Ingenuity, Meta-
Core
from Ariadne Genomics. All these tools are
designed for global pathway analysis of complex
data sets. Although these tools make it easier to
perform comprehensive analysis, they are only
as good as the primary data they draw from.
In the following section, we review two of
the most commonly used methods for
protein
e
protein interaction discovery and high-
light some of their shortcomings that directly
affect the quality and the usefulness in biomarker
identi
from GeneGo, and Pathway Studio
cation.
COMMON EXPERIMENTAL
METHODOLOGIES TO
INTERROGATE
PROTEINePROTEIN
INTERACTIONS
Although strict reporting policies might help
generate high-quality data in the databases, the
impedes interactions between
Search WWH ::
Custom Search