Through many web searches, and much trial and error, I finally got my system up and running and working how it’s suppose to. Mostly.

Setting up Apache:

This was fairly easy. You can either go to the sharing screen in System Preferences and check Web Sharing:

Or you can start it via the command line:

> sudo apachect1 start

Working with PHP:

PHP5 comes pre-installed on Leopard, but it’s not fully functional with Apache. You have to go into the Apache configuration file and tell it to use PHP.
Open up /etc/apache2/httpd.conf and uncomment the line (remove the #):

#LoadModule php5_module libexec/apache2/libphp5.so

and save the changes.

Installing and running MySQL:

The installation of MySQL went easily and quickly, as most things do on a Mac. However, there were a few problems.

First, the latest version of MySQL for Mac (5.1) doesn’t install in the same place as the system expects it to. Instead of installing in /usr/bin like the system expects, the executables install in /usr/local/mysql/bin. This means you have to change your environment to point to this location.

$PATH:/usr/local/mysql/bin

The MySQL starter package caused me the most troubles. I couldn’t seem to get MySQL to start up when I booted my computer - even the GUI in System Preferences wouldn’t start the database package.

The problem turned out to be the location where the MySQLCOM folder was installed. The default location was /usr/local but Leopard needs it to be in /Library/StartupItems for it to work.

So you have to move the folder into the right place:

> sudo cp -r /usr/local/MySQLCOM /Library/StartupItems

That should make it possible for MySQL to start up automatically every time you turn your computer on.

MySQL installation has a habit of putting files in the wrong places on Leopard, the most annoying (so far) being the location of the mysql.sock file.

Trying to integrate MySQL and PHP was a trial. PHP is looking for mysql.sock in /var/mysql/mysql.sock, but MySQL installed it in the old location /tmp/mysql.sock.

This is easily fixed, once you know what the problem is.

There are two ways to do this. The easy way:

> sudo mkdir /var/mysql
> sudo chown _mysql /var/mysql
> sudo ln -s /tmp/mysql.sock /var/mysql/mysql.sock

This is the best way - and easiest - way to make it all work, especially if you want to run the DBI perl module on your system. (DBI still looks for /tmp/mysql.sock.)

Or the harder way (but is probably better from a programming stand-point):

First, create a my.cnf file in /etc and add in:

[client]
socket = /var/mysql/mysql.sock
[mysqld]
socket = /var/mysql/mysql.sock
to /etc/my.cnf

Then, copy the socket file over to the proper directory for PHP to access. Also, create a symbolic link in the /tmp directory to the /var/mysql/mysql.sock so that other programs that depend on that file being in /tmp (ie. perl DBI::mysql) will still be able to function.

> sudo mkdir /var/mysql
>sudo chown _mysql /var/mysql
> sudo mv /tmp/mysql.sock /var/mysql/mysql.sock
> sudo ln -s /var/mysql/mysql.sock /tmp/mysql.sock

This may end up causing problems if you’re using the GUI aspect of MySQL, but since I’m not, I didn’t bother to fix that.

Apache needs to be restarted for the changes to take effect:

> sudo apachect1 restart

PHP should work after this.

You can find out how to solve the MySQL GUI issue, as well as setting up virtual hosts on Apache from this blog post.

This solution, however, also causes problems when using the DBI perl module. As I stated above, DBI looks for /tmp/mysql.sock and if it doesn’t find it (as /tmp is emptied upon computer shutdown), then it returns an error.

I started off with the harder, more complex way, but ended up switching to the first configuration I mentioned, as it’s easier and causes less problems.

…Asia’s oldest bioinformatics organisation set up in 1998. (Bioinformatics research in the Asia Pacific: a 2007 update)

I simply find it worthy of note that I work and study in a field whose ‘old and venerable’ organizations are between 10-20 years old.

Makes you realize how far we’ve come in such a short time.

One of today’s posts feature Eliza Gauger, the designer of the io9 logo - which is very unique, trendy and cool. She talks about a few different things, including how she became converted to the scifi monsters of the world. However, in my opinion, the best part of the entire interview is when she starts talking about what she hopes to never see again in scifi art, one of which is “hot people”.

Look at Aliens. Or any Cameron film from that era. If he had attractive women in his movies, and he did, they weren’t “women” in the way that movies define women: harpies, hags, or idiots. Scifi ditto. Ripley was not wearing any fucking mascara. She was a CHARACTER, she wasn’t a GIRL. Ditto for everyone else. They were people before they were badasses, or killers, or idiots. This is not a luxury in sci fi. It is a necessity. Cookie cutter characters are unbelievable, and in a wider context of unbelievable things happening (aliens, lasers, spaceships, all imaginary), it is vitally important for as much of the rest of the package to be well-developed and believable.

These are the types of people I hope mold and promote scifi away from the glamour and shiny-ness of Hollywood. Scifi movies - and really, any movie that has a modicum of science in it - are already on a downslide credibility-wise, that having perfect people starring as humble scientists doesn’t really endear a lot of people. So, putting Denise Richards in a lab coat and glasses does not make her any more credible as a nuclear scientist, than if you’d cast some less-than-perfect actress. In fact, it might even detract from the movie (I know I rolled my eyes).

However, bad movie science is for another, and longer, post.

Click on the image to go directly there.

Needless to say, it’s been changed.

When Munroe took over as chief two years ago, his department was drowning in crime and data. Police had a mass of data from 911 calls and crime reports; what they didn’t have was a way to connect the dots and see a pattern of behaviour.

Using some sophisticated software and hardware they started overlaying crime reports with other data, such as weather, traffic, sports events and paydays for large employers. The data was analyzed three times a day and something interesting emerged: Robberies spiked on paydays near cheque cashing storefronts in specific neighbourhoods. Other clusters also became apparent, and pretty soon police were deploying resources in advance and predicting where crime was most likely to occur.

Link - via CBC

I had one sequence map itself to an exon and a 5′UTR - on the same transcript. This caused me some concern (and an afternoon of frustration), because it seemed that according to the Ensembl database, this one position is both an exon and a 5′UTR. I thought I had done something wrong in my code, such as calculating the start and end position of the UTRs wrong or pulling out the wrong starting position for the start of translation.

Seems that wasn’t the problem. What it turned out to be was partially my fault, but also partially the fault of the Ensembl gene annotation.

First, my protocol for determining locations was done in two steps. A - I retrieved all the exons (start and end positions) that were associated with a certain transcript, and used that to determine whether my tag was located within an exon, if it overlapped one of the ends of the exon, or was part of an intron. B - I pulled out the start and end of translation, which is given by the first exon and the last exon in the translated region (then pull the start and end positions from those). Anything before the start of translation was the 5′UTR and anything after then end of translation was the 3′UTR.

This is where the problem cropped up.

You see, I assumed that the translation encompassed all the exons for that transcript, and that all exons were translated. I was wrong. There are exons that are not translated (ie. UTR), but are still considered exons. (This is where ‘talking to others’ comes in handy.)

I found that out of the 2800+ transcripts I was looking at, over 1100+ of them had an exon associated with the 5′UTR. It didn’t get translated, but it was considered an exon, all the same.

Solution: Determine whether you want to call it an exon or 5′UTR. In this case, we called it a 5′UTR. So, whenever a tag mapped to the 5′UTR of a transcript, I made sure it wasn’t mapped to an exon, as well. It was just a matter of re-arranging my code so that it looked for the 5′UTR first, and then added a flag to prevent it from evaluating it as an exon/intron.

Lesson learned: Always have an idea of what you’re data should look like, and actually ‘look through it’. Take out some random test cases - or some odd looking results - and backtrack your way through the code. Or verify these test cases through database searches or by-hand calculations. If you know what you’re suppose to get, it’s easier to figure out where the problem is, when compared to what you actually get.

Note to self: Learn the Ensembl API!!

_”The primary goal of this work is to convert genome-encoded protein sequences into musical notes in order to hear auditory protein patterns.”_

This is an interesting - and unique - approach to analyzing protein sequences. They’ve taken many things into consideration, including codon distribution and amino acid frequency. They’ve developped rules for rhythm, range, dynamics and note lengths. Unlike previous efforts which have tried to directly translate an amino acid to a musical note, this approach uses chords and musical rules to generate these compositions.

Previous efforts to translate amino acids into notes didn’t work too well, as there seemed to be quite a few sudden shifts in range. Essentially, they sounded like a three year old banging on the piano. This approach used by Takahashi and Miller managed to reduce it down to 13 base notes _”…according to hydrophobicity and pairing of similar amino acids.”_ I don’t have a history in music or musical theory beyond three years as a clarinet player in elementary school, so I have little to no idea what’s going on when they start talking about ‘three-note chords’, ‘root position’ and ‘first inversion chord’. (Anyone care to evaluate this for me?) The compositions are not simply ‘note-note-note’ such as you would find in a one-on-one translation of amino acids to notes. Instead, they developped chords, which really fill out the composition and make it more auditorily (sp?) pleasing. Then, the rhythm was added, based on codon distribution among a set of 1000 codons. This approach provided the pacing for the piece, making it much more musically pleasing.

Listening to this, you can hear the repeats and other specific domains in a protein. Huntintin gets slightly annoying after a few seconds, but what can you expect from a protein with that many tandem repeats? After listening to a few samples, you realize that there is something off with their algorithm, because you start hearing the same sequence of notes in the same pattern, repeating through different proteins. This could be a biproduct of their algorithm, or it could just be that I’m hearing similar domains in each of the proteins. I’m leaning more towards the second explanation, because after a while your mind starts to associate the tones to one another. They only have 13 base notes, after all.

Personally, my favourite is the horse hemoglobin.

But blogging isn’t just about you getting involved in the field. It’s also about distribution of knowledge, transparency in science and global communications, which all lead to a better you. Having a conversation over a blog post can be a great way to develop your reasoning and researching skills. A lot of people starting out - especially those who haven’t gone through a thesis defence - have a hard time holding a reasonable ‘on topic’ conversation with their colleagues/peers. In most cases, it’s not because of lack of knowledge, but because of an ability to express your ideas in a concise and reasoning manner. A good ‘on topic’ conversation between two people in the same/similar field will be intelligent, well articulated and will make each participant consider new ideas or perspectives. (This type of conversation or debate is not solely isolated to the science field, but that’s what I’ll concentrate on.)

Communicating with bloggers through comments and posts on their blog or your own (if you have one), is a way to have this same type of conversation, but without the necessity of replying immediately. If you’re like me, five minutes after someone expresses some statement or idea, you come up with a response, but by then the conversation has passed on or ended. On the web, you have a chance to reason through one person’s perspective and come up with counter arguments or alternative perspectives. You don’t have the immediate pressure of face-to-face dialogue, so you can take a breath, close your eyes and determine what you want to say and how to say it most effectively. You’ll find, over time, that your ability to ‘think on your feet’ becomes much improved.

You also have a chance to fact-check, which is always a plus when trying to convey a perspective or argument.

Leaving the conversation aspect of blogging behind, bloggers are at their core, disseminators of information. Quite honestly, none of us have time to scour the web every day for every bit of information relating to our field - well, not and do our jobs, as well. By browsing through several different blogs, each focusing on different aspects, you can keep updated on a lot of the newer technologies/techniques/breakthroughs that are happening in your field.

You can find blogs on almost any subject, but bioinformatics-related blogs - I’ve found - tend to be some of the most diverse. They range from programming tips/languages to science issues to ethical debates to political/religious topics to the latest trends in tech and abroad.

And they’re well written. I think that’s one of the most valuable aspects of blogs nowadays - the concise and even-level tone of blogs. There are thousands of great writers out there, but until the advent of the internet and blogs, they’ve been mostly confined to published works. These people then tend to become more about the writing, and less about the science they’re writing about. Blogging for bioinformaticians, as well as for many other people in many diverse fields, is done as a side project. It’s something they put thought and effort into, without the expectation of monetary reward (usually) or having to fit into a set mold (word count, views, topics, etc…). The net is not simply about net-speak and social networking sites, it’s about expressing opinions in a well thought out manner that conveys the ideas and arguments of the writer in a reasonable way. Also, having a well-written blog post garners more attention and credibility than one filled with grammatical errors, no punctuation or capitalization and overflowing with net speak. If you want to be taken seriously on the internet, you have to sound like a professional. Often, people won’t know who you are or what you do for a living or even how you act, so your words need to convey that you are intelligent, open minded and well-read, and that your opinions matter just as much as everyone else’s. Your words are all they have available to make an impression about, so don’t sound like a whiny twelve year old.

Blogging is also a good way to keep in touch with colleagues and friends as you move through different positions and jobs. It’s a more passive way to keep in touch than email, but just as powerful. In this field, an extended network of friends and/or colleagues are as good as gold (see above re: collaborations).

But reading and writing blog posts is not the end. Read the comments, too. Lots of good ideas are put forth in blog comments, including links to other material. I find a lot of my reference links that way.

Blogging has opened up a whole new world for the scientific community. Editorials and reviews are no longer the purview of published journals alone, and meaningful, intelligent discussions can occur outside these journals’ discussion boards. As the internet grows, the world becomes smaller; people are able to put forth new ideas, or discuss old ones, advancing personal and professional alike. There’s a transparency developing online, in just about every aspect of life. There are no ‘big secrets’ anymore, because there is a repository of knowledge on the internet, just waiting to be found and disseminated. The scientific community included.