Table of Contents

1. Introduction (Terminology)
2. Structure
3. Pearls
1. Getting started with free energy
2. Everything else
1. Pulling genes and taking names
2. Making wait times and Travel.mat for any bacteria
3. Converting between amino acids and mRNA
4. Unities
1. A review of ribosomal translation
2. A brief tour of the internals
3. unity and megaunity
4. Everything else incubating
5. Loose ends incubating
6. Questions that I think are or should be asked frequently, or QAAF incubating

Introduction

(For a theoretical treatment of this project, please read our research paper.) Biologists can tell how well Escherichia coli translates a certain mRNA strand with various experimental methods. Is there a way to replace that process or parts of it with a computer?

Terminology

A gene file refers to a file in the FASTA format. However, our model simply ignores the FASTA header line that describes the gene. As a result, you can also input a file that is purely the sequence of bases that make up a gene. For simplicity's sake, the de facto file extension we've adopted for this relaxed FASTA version is just .txt, which is how we name all the genes we publish and all the genes to which we shall refer in this manual.

Structure

Our model is divided into two sections: Matlab and Perl. The entirety of the Perl section lies in the pearls folder. Each file has (or should have) complete POD documentation. That is, if you type the program name at the command line without any arguments, the help text generated by Pod::Usage should show up. Try it out with scan_brightly.pl. On the Matlab side, the documentation is written in comment form for each file. Each Matlab file corresponds to one major, documented (hopefully) function. To access it, use help from Matlab's interactive environment. Try it out with help unity.m.

The Matlab files within the unities folder are thematically united by the purpose of data collection. Each file collects a specific statistic (for example, deviation from the zero reading frame) usually over a number of samples and many, many genes. As a result, many unities can take a directory or a gene file as an argument.

The files within the God 2.0 and the thrushbaby directories are part of now-defunct algorithms. They are further described in the "Algorithms" section. The files within util are utility functions used by some Matlab files. You might find them useful in constructing new tools based on our model. They're friendly folks with open hearts and beautiful faces.

Pearls

Getting started with free energies

Free as in C memory, not free as in you burnt your thighs with our coffee and now we have a settlement. The primary Perl program exposed to users is the scan_brightly.pl file, which is an interface to our Kidnap Perl module. The command is simple: scan_brightly.pl prfB.txt and a list of free energies at each codon spills into your command prompt. Usually you will never have to type this because the list is useless without the Matlab portion of our model. However, let this not mislead you to believe that it's not important. Our entire model rests upon these free energies. They are our cornerstone and, most importantly, our friends.

For a more in-depth treatment of all the pearls, see the Pearls section of the manual.

Unities

M> unity('rpoS.txt');
M> unity('prfB.txt', [25]);
M> megaunity('prfB.txt', [25]);
M> megaunity('prfB.txt', [25], 100);
M> megaunity('rpoS.txt', [], 100, 'graph');

Ribosomes and me

Free energies provide only so much sustenance to an overworked laboratory molecular biologist. By taking a derivative, we arrive at force of the 16S tail acting on the ribosome. I suppose this is as good a time as any to review translation in prokaryotes.

The cell reads DNA, which I'm mandated to say is the genetic code of life, and transcribes it into mRNA. A ribosome latches on to the starting codon of mRNA. tRNAs, little bugs, attempt to pair with the codon the ribosome is latched onto. The first one to be correct wins. (We call the length of time between latching and matching wait time.) The ribosome, happy, takes off the amino acid on the tRNA, sticks it to what it already has, kicks the tRNA out, and moves on. This continues until the ribosome sees a stop codon.

From a 1987 Weiss paper, we know the 16S tail can nudge the ribosome into skipping a base pair (frameshifting). When? We calculated the force earlier. If we run the force through the stochastic model, which is described in detail in our research paper, we get a vector containing all the changes in displacement made at each base. If we sum all those changes, we get total displacement from the zero reading frame at each base. As a side effect, the model also can output where the ribosome frameshifted. For most genes, correctly frameshifting means not doing so at all. For prfB, however, an intentional shift exists at the 25th codon, which is an uga.

Unity and Megaunity

unity.m and megaunity.m measure the deviation from the zero reading frame or the probability of frameshifting correctly. (Which one it chooses depends on the Config.yield assignment in config.m.) unity.m is the simpler one. You pass it a file and it graphs the displacement as well as outputting the yield and frameshift locations like so:

>> unity('rpoS.txt')
> guu 76; ucu 89; aaa 103; gcc 105; cgu 111; aag 132; caa 160; acu 163; cgu 165; cga 179; gca 181; uug 184; cau 190; gag 196; gac 222;
< auu 77; uaa 102; uug 104; gca 109; uag 130; uga 158; gua 162; uuc 164; cga 167; ccg 180; gug 182; ugg 188; aag 195; uag 221;

Yield: 1.48858

"> guu 76" means a forward frameshift occurred at codon 76, which is guu. "< auu 77" means a backward frameshift occurred at codon 77, which is an auu. If you're running prfB or any other gene that has an intentional frameshift, you also pass it a vector telling the model where those occur. For example, unity('prfB.txt', 25). Remember, 25 is the same thing as [25] in Matlab. Matlab is genius like that. And one day I'll discover the gene that makes bacteria happy. Maybe that gene will frameshift in three places: unity('happy.txt', [12, 25, 81]). Won't that be great?

megaunity.m runs exactly as if you ran unity.m forever. It averages the yields over the run. So if you just want to run prfB forever, megaunity('prfB.txt', [25]). And if you wanted it to stop after 100 times, megaunity('prfB.txt', [25], 100). And if you wanted to look at the graphs, which are hidden by default, megaunity('prfB.txt', [25], 100, 'graph').

For a continued treatment of the Unities, see the Unities section of the manual.