The Mistakes of The World's Worst Lab Tech

How am I not fired yet?

[ipipettethings]

Who the fuck decides how fast to spin things?

One thing that has always and continues to bother me is… who figures out how fast to spin things and for how long?  Where did this magical number come from?  Do we eventually reach a point where we know exactly how hard and how long to spin whatever type of cell or cellular product we want, to the point where it's compact… but not too compact?

And why does EVERY PROTOCOL have "xg" for the speed??!  I want RPM damnit!

Don't make the mistake of thinking rpm's correlate exactly with xG.  Sometimes it does… but then you're lucky.  Why can't the fucking centrifuge have a xG setting, huh?  Well… some do.  It's called RCF.

"Relative Centrifugal Force" is the amount of acceleration you're applying to your sample, relative to the standard acceleration due to gravity, g.  You can calculate it with a formula… but who gives a shit.  There's online converters for that. 

The important thing to remember is:

1. RCF and G are interchangeable.  If it says spin it down at 400xg, go ahead and switch to rcf mode and put in 400.

2.  Use your rotors radius to calculate RPM from xG if your centrifuge doesn't have an rcf option.  Many online converts exist if you're lazy (that often have the precise radius of whatever rotor you're using..)  It's nice to use the formula in an excel sheet and have the conversion for every 50xg increments on top of the centrifuge.

So.. if you're starting to realize that 1,200 rpm on a little Eppendorf centrifuge isn't even close to the same acceleration as 1,200rpm on a large Beckman-Coulter centrifuge… well then good fucking job.  You're smarter than I was.

[Don't lyse precious samples because you thought you were spinning down at 800rpm but instead were spinning at 800xg.  If you try to wash, chances are you just washed the majority of your cell lysate out after re-suspending.  Sucks bro.

Conversely… don't spin down at 800rpm when you're supposed to spin down at 800xg.  If you're not a pussy, you'll use a vacuum to aspirate… and then say bye-bye to your precious cell pellet :( ]

[ipipettethings]

Hey. Thinking of becoming a lab tech after my undergrad. I thought it would be a great position to have while really deciding/ applying to grad programs. What was your major? How did you end up as a lab tech?

I was in the same boat.  

I don't see how people can make a decision about the next four to eight years of their lives without actually seeing what being a physician / post doc is like!

I wasn't sure what I wanted to do, and since lab jobs don't pay shit in undergrad (at least here) and I couldn't afford to volunteer, I needed the experience anyways.  You really do need some sort of lab experience to go anywhere.

I majored in molecular and cell bio, but it doesn't matter too much.  It doesn't really matter what your GPA is either.  

What matters is two things:  

That you make an impression on someone who can give you a good recommendation.  PI's receive hundreds of applications for these types of jobs, and in fact the job market for them is so shitty that long-time veterans are getting laid off and applying for the same pathetic entry-level positions that we are.  A recommendation goes a longggg way.  In fact, I didn't get a single response to a job I applied to on my own, except boiler plate rejections from biotech companies.  But the first PI that one of my professors forwarded my resume to called me the next day to setup a meeting.  

It's not what you know, it's who you know.  That old saying sucks a fat one since it ruins the idea of a meritocracy, but after meeting a lot of people in this field, I think I get it.  Science nerds are fucking weird.  If a person can make such an impression on a professor/PI/someoneimportant that they go out of their way to advertise you to their colleagues, then that must mean you have *some* social skills.  A really bright person isn't useful to have in a lab if he can't communicate well with others, handle criticism, work in groups, or if nobody likes to interact with them.

And number two: you need experience.  Volunteer and take out some more loans if you need to, but undergraduate lab experience helps a lot.  I somehow got away with only class lab experience, but then again this blog IS about all my fuck-ups as the worlds worst lab technician.

[ipipettethings]

Metric Probs

As an American scientist, I still can't convert from imperial to metric for shit.

Know that 1 US gallon is about 3.8L, or 4L if you're really lazy.

Maybe it's because I just sent out a package that day (you know, FedEx tells you to estimate 2 lbs per Kg), but 1 Gallon does NOT equal 2 Liters.  So picture this… it's my fourth day in lab and I don't know how to do a god damn thing, especially order shit.  I'm worried because I can't even do simple dilution math without getting all pissy, but it's my job to make the ethanol dilutions, the job nobody wants.

Alright, well how big is our container?  10 Liters.  How much ethanol should I order?  Fuck it, how about 40 Liters?  Having three refills on hand will be nice right.  Okay… hmm so that's about 20 Gallons.  NOPE. 

Now I have to store 20 Gallons worth of pure ethanol, and there's no way in hell this will fit in a "fire safety" cabinet.  Nope.  Every single bay in our lab and neighboring labs has a metric fuckton of ethanol below each of their sinks.  If there's ever a fire in this bitch, we're all going to die in an accelerant fueled inferno.

And the sad part is there's no reason to not know this shit, hell you see the conversion every time you piss into an American Standard urinal.  (are there actually other brands?)

[ipipettethings]

Math is hard. Dilutions suck.

What's the most degrading bitchwork that you can possibly do as a lab tech? - making ethanol and bleach dilutions.  (well, besides autoclaving)

So you're making a 70% ethanol dilution huh.  Well that part's easy, just add 7 parts ethanol and 3 parts water right, smartass?  It's even easier when you have markings on the fat jug you pour it into…

Except if you let it drain to empty, you look like a bad lab tech.  Now your 10L container is slightly more than 20% full, and some genius bought fucking 190 proof ethanol instead of 200, so have fun with that.

In this case, you'd probably want to make 7.5 liters total of 70% ethanol to add into what's left. To be honest you can probably just eyeball it and the universe would still be the same, but if you want to be anal... 

7.5L = 0.95x/0.7  where x is the total volume of ethanol you're going to add.  

Since that hurts my eyes, just multiply your target volume by 70% to get the amount of ethanol you need to add, and divide by 95% since it's 190 and not 200 proof ethanol. (proof/2=% for you non-drinking fags)  Subtract the volume of ethanol from 7.5 to see how much water you need.

= ~5.53L of ethanol + 1.97L ultra-pure H20 (e.g. MilliQ water) added to your jug.

Note:  For bleach dilutions, just dump that shit out.  Bleach dilutions lose their effectiveness relatively quickly.  You should probably change it at least once a month.

[ipipettethings]

Coding vs Non-coding Strand

So remember in every single powerpoint slide and every single book figure that you've seen about DNA transcription, there's always that one strand on the top that polymerases work off of.  That strand is the shit, the big guy who always gets the final say.  That other strand is just it's complementary bitch that stabilizes it and gets no love, except during cell division… and then it gets all those Okazaki fragments, gross.

So when you hear the words "coding strand", what strand do you immediately think of?  I'll tell you what I think of, and that's the definitive, authoritative strand that RNA polymerases work off of while the bitch strand hangs there.  Except it's not.

The coding strand is the bottom strand because it's sequence is identical to the messenger RNA strand sequence (with the substitute of uracil for thymine of course).  

This may seem easy to remember as a concept, but when you're looking at a shit ton of AAATAGCCGGGCCTA's all over the screen and start to wonder hmm is this the strand that makes proteins or the other one? Should I reverse it?  Should I choose the complementary base pair sequence?  Should I reverse it AND choose the complementary base pair sequence??

That shit is confusing.  Remember, coding strand = mRNA.  It's always read 5' to 3'.  Don't look like a retard in front of your PI.

[ipipettethings]

Latex vs Nitrile Gloves

So for some reason the simplest things bother me.  Like why did you choose to wear THAT brand of gloves, in THAT color, made out of THAT material.  Nobody seems to give it a second thought, but if you look around, you'll start to go crazy.  You got the plain latex and nitrile gloves, powdered and non-powdered gloves, gloves with criss-cross meshes, thick gloves, thin gloves, textured gloves, dark blue gloves, light blue gloves, purple gloves, violet gloves, and gloves with fucking aloe vera already inside them to keep your hands nice and moisturized.  Jesus Christ.

So if you're like me, then you immediately stop and ask somebody wearing a colored glove you haven't tried before if you can steal a few.  Then you find out that some are softer, more tight-forming to your hands and others are thicker and better textured.  What to choose, what to choose...

Let's go back to Latex vs Nitrile to make it simple.

Why choose latex?  Well it's extremely form-fitting to your hand and offers unmatched finger dexterity.  The latex itself is pretty sticky so it gives you good grip, but you should usually get the un-powdered kind as the corn starch can fuck up your science.

Why you shouldn't choose latex:  I don't know about you, but I never sit somewhere and just do ONE thing.  I have to constantly take my gloves off and put them back on.  Since we're poor… that means using the same gloves.  And this is ridiculously hard to do with unpowdered latex.  It's also harder to see holes in your gloves with latex if you accidentally poke yourself with a needle… or have one of those black devil mice bite you.  AND the biggest one.. latex smells like shit.

Why choose nitrile?  Well apparently nowadays you can get nitrile gloves in any fucking texture, color, form-fittingness factor, and thickness.  I prefer the dark blue nitrile gloves from Microflex because they have medium thickness, a decent texture on the fingertips, a good enough fit that gives me the dexterity I need for most tasks, but most importantly they don't adhere to my fucking skin like superglue so I can actually take them off pretty easily.  And.. they don't smell like a dentist's office.

Sooo, all that aside, my labmate put it best.  You can develop allergies to both materials.. but would you rather develop allergies to latex, or would you rather develop them to nitrile?  I was quite ambivalent, until she reminded me something that could have saved my life:

Nitrile condoms are thick and sensation-less pieces of shit.

Latex condoms are muy bueno.. well.. if you have to use one :(

P.S. you can peel the gloves off, turn them back inside and blow into them.  This still grosses me out because you have to touch at least the top of one glove with the other glove.

P.P.S. I sweat a lot.  Latex gloves turn brown and stain when I wear them.  Sweating inside gloves increases chances of developing allergies (or so I'm told..)

[ipipettethings]

Proper Point Mutation Notation

Shit they don't teach you in class, part 1:

We often get patients, we often sequence their genes, and it often happens that they have shit wrong with them (they all have cancer so… big surprise).  So say a patient has a single point mutation in NRAS, how do you document it?

I'm just a lab tech.  I didn't design these primers.  I didn't choose what exons we sequence.  Hell I didn't even put in the order to get these kids genotyped.  A  sequence file magically appears in a network folder that I check every-so-often.

So what's the first thing you do?

1.  Make sure the sequence viewing program (e.g. FinchTV) called the bases correctly.  If your signals are weak (like in the low hundreds) and you have a lot of noise, then clean that shit up better before you send it to them next time.  

2. If there's no abnormal spaces, then check for double-peaks.  If you see two peaks at one point, then the patient has a heterozygous point mutation.

3.  If everything looks good, let's compare it to the reference sequence.  Everything you've read has suggested you do a BLAST alignment at this point, right?  Nah, fuck that.  All the cool kids are using BLAT now (genome.ucsc.edu) even though it was developed by a lesser UC known as UC Santa Cruz (suck it banana slugs).  So do an alignment and check for differences.

And of course, remember the beginning and end of the chromatogram will look like absolute shit.  This is normal.  I believe it not only has to do with the fidelity of the polymerase as it tries not to fall off the wagon after ~1000bp, but also because the gel has such bad separation between bases towards the top it's unreadable.  (Using Sanger sequencing here people, because we're poor)

So I found a point mutation.  My PI asks where it's at.  Okay no problem.  It's some random ass number of bases past where the sequence reader started :)

No.  So go back to the UCSC genome thing and search for your gene.  Click on your gene in the genome browser and go down to Genomic Sequence under "Sequence and Links to tools and Databases".  (There's got to be a more direct way… but god forbid you even try to do this in ncbi BLAST).  So from here you have the option to show the UTR's, introns, exons, etc.  Exons here are capitalized to easily distinguish them, but apparently UTR's also count as fucking exons.  And this is important.

You might be tempted to start counting from the first bold stretch of letters, but this is the 5' UTR.  Bad.  Go down one more and this is exon 1.  Now… don't do what I did and just start counting from the beginning of exon 1, you know, like you would think makes sense.  

Oh no, the beginning of exon 1 is just some random ass splice site where the UTR continues.  Start counting from the first start codon, which is methionine in eukaryotes (for the most part).  

Sooo, you might remember that the start codon is AUG.  Well dammit ctrl+F doesn't find an AUG.  Hmm duh because BLAT isn't showing RNA, it's DNA.  Hmm so I should look for a TAC?  No.. wait CAT?  No.. fuck.

Remember, BLAST and BLAT searches will show you the coding strand (unless it tells you otherwise).  This is the strand that is identical to the mRNA, so all you have to do is substitute Thymine for Uracil.  The start codon is ATG.  The first A of that ATG is nucleotide number 1.  From there you can start counting.

(Other exons are super easy, because the beginning of the exon is the first nucleotide.  Remember they all get spliced together, and the ribosome is just gonna translate everything it comes across after the start site so there better not be random shit in there)

P.S. I think I'll write more on how to actually read chromatograms properly.

P.P.S. If you came looking for the full mutation notation, it's very poorly written by the standards committee that makes them up… see http://www.hgvs.org/mutnomen/recs-DNA.html#DNA