My apologies for possibly a naive question. I'm a statistics type and have been asked to assist with analysis of metabolic pathways (and networks). However, I would like to have at least a layman's understanding of how metabolic pathways/networks relate to DNA sequences. Been doing the Google for a few days and, given the results, it's clear I am not formulating the query correctly since nothing seems to be clicking.

Something that helps me understand the mechanics behind going from a string of T,A,G, and C's to a network of substrates,enzymes, and products.

Thanks in advance for your patience.

  • 2
    $\begingroup$ Welcome to Biology.SE, I fear this question might be too broad to be reasonably answered here. You might want to just start by reading an entry-level molecular genetics textbook. Any chance you can make your question about specific things you have a problem understanding? $\endgroup$ Apr 12, 2012 at 21:14
  • $\begingroup$ Yeah, I was afraid of that. I was looking for a very introductory level bit of material(s) and was hoping that someone had put something together. I'll understand if you close the question. $\endgroup$
    – Aengus
    Apr 13, 2012 at 3:18
  • $\begingroup$ Do you mean the connection as it occurs in the cell, or how to model it (like, how to predict a function for a particular gene)? $\endgroup$ Apr 13, 2012 at 5:57
  • $\begingroup$ The long Question is how to predict a function. To try to be more specific, I've tracked down the paper by A. M. Feist, M. J. Herrgård, I. Thiele, J. L. Reed, and B. Ø. Palsson, “Reconstruction of biochemical networks in microorganisms,” Nat Rev Micro, Feb. 2009. I'm trying to dig into Figure 1. $\endgroup$
    – Aengus
    Apr 13, 2012 at 14:24

1 Answer 1


The question is not too broad, just involves a fair bit of work to both do the research and compose a response. I'll do the latter, but in brief.

The strings of the four nucleotides encode genes. Sometimes these genes are broken into protein-coding portions (exons) and sometimes intervening, non-coding regions (introns). Bacteria, for example do not have introns within genes while most human genes do. The nucleotides in the protein-coding portion of the gene do just that - give instructions on how to synthesize the protein by directing which tRNAs the ribosome uses to translate the genetic code into a protein chain.

A useful phrase is protein sequence dictates protein structure and that structure dictates function. So, the proteins do things and many act as enzymes or modulators of enzyme activity. An enzyme catalyzes a biochemical reaction, lowering the activation energy to go from reactant(s) to product(s). One enzyme can be thought of as one unit within a metabolic network. A modulator might activate or deactivate an existing enzyme. A protein kinase is a good example of a modulator.

Next, string some of those enzymes together to build a pathway. A common example is glycolysis. Think of this as a chain where the output or product of enzyme A serves as input to enzyme B, whose product is the input to enzyme C, and so on. These pathways are two-way, but it is exceedingly difficult to travel from the final product all the way up such a pathway to the initial input within the confines of a biological system. Thus, the pathways in practical terms are thought of as one-way.

However, the pathways do not function in isolation. The final product may serve as input to one, two or more other pathways. Your pathway of choice may begin with the output of another process. Furthermore, pathways can branch: Glycolysis can proceed to glyceraldehyde 3-phosphate and pyruvate, or directly after glucose 6-phosphate formation can be shunted to the pentose phosphate pathway to yield reducing power and 5-carbon sugars (to be used in ATP, CoA, FAD, NAD+, DNA and RNA). Now, pathways leading to other pathways and branches and such make a network.

You and your colleagues might want to know if genes within a metabolic network are over- or under-represented in a test compared to control. Or, genes for a given network are not present in one species compared to several others. Or, one could estimate the flux through such a network given a set of measures.

Peruse examples at both KEGG and Reactome, expecially at KEGG where you can view pathways pertinent to your organism(s) of choice.

  • $\begingroup$ Many thanks for getting this started. I understand (or least have an intuition about) everything after "... and that structure dictates function"; and I've wandered through the KEGG database. So maybe the question is better phrased: How is the protein structure determined/predicted from the protein-coding portion? (I think this is the 'reconstruction' part?) Thanks again for you patience. (I should add that I'm not immune to doing my own digging; a few pointers in the right direction would be appreciated.) $\endgroup$
    – Aengus
    Apr 13, 2012 at 14:38
  • $\begingroup$ In general, assignments of function are based on sequence similarity (tools like BLAST) - comparing protein sequences to those with known or demonstrated function - or identification of functional motifs as in the Pfam, SMART, INTERPRO and others (ca.expasy.org/tools). $\endgroup$ Apr 13, 2012 at 15:37
  • $\begingroup$ Still a few open questions, but I think that gives me enough to start doing more focused background work. Thanks! $\endgroup$
    – Aengus
    Apr 16, 2012 at 15:47

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .