Biomedical Engineering Reference
In-Depth Information
Viewing the Central Dogma from the perspective of Information Theory enables researchers to apply
computer-based numerical techniques to models and evaluate the underlying processes. Evaluating
the Central Dogma in terms of information flow also brings metrics into play—such as the
effectiveness of the underlying data archiving capacity, the effectiveness of the information flow
process, error rate, the degree of data compression, and the degree of uncertainty in the data
translation process.
Each of these concepts is related to computer science information theory technologies and
approaches. For example, DNA functions as a data archive and, as such, can be evaluated as any
other information archive. There are issues of information capacity, how data are represented within
the DNA molecule, whether there is provision for automatic error correction, the longevity of the
information, the various sources of error at different points in the system, and how the information
embedded in DNA is shared with other systems.
Even though there are apparently only about 40,000 coding genes in the human genome, a typical
human DNA strand would extend several feet if uncoiled. The physical compression of the nucleotide
sequences into tightly coiled chromosomes has parallels in the digital processing world where there is
a constant tradeoff between storage capacity and access speed.
As a final example, consider that there is a degree of uncertainty inherent in the communication of
information from DNA to protein. For example, the process of information flow can be analyzed to
model the sources and types of errors (such as mutations) in the flow of information from nucleotide
sequences in DNA to protein in the cytoplasm. In addition to modeling and simulation, resolving or at
least quantifying this uncertainty entails the use of data mining, pattern matching, and various other
forms of statistical analysis.
From Data to Knowledge
In viewing the Central Dogma as an information flow process, it's useful to distinguish between data,
knowledge, and metadata. For our purposes, the following definitions and concepts apply:
Data
are numbers or other identifiers derived from observation, experiment, or calculation.
l
l
Information
is data in context—a collection of data and associated explanations,
interpretations, and other material concerning a particular object, event, or process.
Metadata
is data about the context in which information is used, such as descriptive
summaries and high-level categorization of data and information.
l
In addition to data, information, and metadata, the concepts of knowledge and understanding are
worth noting because these terms often surface in the computer science literature. There's a trend,
for example, to re-label databases as knowledgebases—a term borrowed from the AI research
community.
Knowledge
is a combination of metadata and an awareness of the context in which the
metadata can be successfully applied. In the same context,
understanding
is the personal, human
capacity to render experience intelligible by relating specific knowledge to broad concepts.
Both knowledge and understanding are normally considered uniquely human. For example, a so-
called knowledgebase of protein folding rules may contain contextual information of folding as a
function of the extracellular environment, but the program using the knowledgebase doesn't have an
awareness of when this context applies. Furthermore, one of the major failings of expert
systems—pattern matching programs that use heuristics or IF-THEN rules stored in a database in
order to make decisions—is the inability to fail gracefully when they are used outside of the narrow
domain for which they are designed.
Similarly, although understanding is often touted as the inevitable holy grail of AI research, even with
the current rate of innovation in computer science, it will be well into the middle of the 21st Century
before machines demonstrate understanding. To illustrate how the concepts of data, information, and
metadata compare with those of the Central Dogma in evaluating the human condition, consider that
to a practicing clinician, a reasonable perspective on a disease such as hereditary disease like
