It wouldn't be called research if we knew what we were doing

[Comoroa]

---

Members do not see this ad.

If you have done research, you understand what it feels like to repeatedly bang your head against a wall to try to get some assay to work. For example:

Today I just found out that all of the cloning I've been attempting to do to get this protein expressed will not work because the initial DNA we had was bad. That's 4 weeks of my life I will not get back.

I'm sure other people had similar stories, and I would love to hear them. Feel free to vent about the horrors of bench research!

 

[DoctwoB]

--Tried troubleshooting PCRs for weeks before figuring out that the primer facility sent us faulty primers.

--Tried yeast transformations for a few weeks before finding out that what I thought was ammonium chloride was mislabelled = no growth

-Protein exprssion issues requiring multiple primer swaps, etc.

I feel like I have an answer for the interview question "Why not pure research?"

 

[BigEast55]

A whole semester trying to get the total synthesis of one molecule before failing at the last step.

 

[Tesseract]

Ran a good 16 qRT-PCR gene expression experiments (different genes) before I realized my total RNA samples were contaminated with DNA. Never having done qRT-PCR before, my 48 RNA samples were sent through reverse transcription without ever completing a DNase clean-up protocol. /oops.

 

[Slowpoke]

I feel so lost at lab meetings. Science has its own language that apparently can't be mastered overnight. It would be a lot easier to just swap out big scientific words for simple logical things.

 

[loveoforganic]

Everything I synthesize in organic research comes out as tar that's impossible to purify. F**k imidazolium salts.

 

[Comoroa]

Everything I synthesize in organic research comes out as tar that's impossible to purify. F**k imidazolium salts.

Hey there are guys working in my lab who are having the same problem with their synthesis. Once they get the tar, it is nearly impossible to get out of the glassware. This is why I'm in biochemistry and not organic.

 

[boaz]

15 hours of pipetting for fluorescence assay and data goes in the wrong direction. 😡

Later reevaluate the whole procedure and realize something simple needs to be adjusted. 😡

Another 15 hours of pipetting coming up. 😀

 

[GoSpursGo]

We spent a good 3-4 weeks trying to figure out why we were getting weird sizes of DNA bands on our gels after cutting our plasmid (with known restriction sites in specific places, so we knew what sizes we should get) with a restriction enzyme... before we realized they sent us the wrong plasmid entirely.

 

[brooklynblunder]

not to drink my own kool-aid, but im pretty good at it. jk, but if you spend enough time in the lab i think you will come across really ****ty plasmids, DNAs that just aren't that good and especially terrible antibodies for western.

its really nice to have a good post-doc that has your back when you need it, but lets you hold your own.

research really only moves when you are doing it 6 days a week, culturing your own cells and ****. sometimes, i hate the teamwork aspect of it because some people either move at different paces, or doesn't take any priority over anyone else's projects


wait til you have to do your figures for the paper. everything needs to be in high resolutions, so you could spend days uploading **** across ftp servers

 

[brooklynblunder]

oh, cell contamination - bacterial or fungus - is the absolute worst. it kills any collaborative environment of the lab and people start finger pointing. always takes up 30 minutes of a lab minute in a really serious mood. people stop sharing medias and you have to make up your own, which fills up the 4C really quickly. and then you run through sterile filters like whitney does pipes

 

[loveoforganic]

Hey there are guys working in my lab who are having the same problem with their synthesis. Once they get the tar, it is nearly impossible to get out of the glassware. This is why I'm in biochemistry and not organic.

Getting the tar out of glassware is what chromic acid, base bath, and piranha solution are for 😀 The trick is getting your compound out the tar!

 

[lorenzomicron]

Getting the tar out of glassware is what chromic acid, base bath, and piranha solution are for 😀 The trick is getting your compound out the tar!

Worked for 7 weeks on Signal Analysis, trying to figure out why my mathematics and programs wouldn't work when applied to certain signals, though perfect in theory, only to be told that it all depended on first extracting the mean from the signal because the signal was not time-invariant. damn. 4 weeks wasted because the stuff i figured out in the first 3 weeks was correct after all...

 

[insane]

working on a biophysics model for 6 weeks before realizing that you misread the rules at the beginning.

 

[Morsetlis]

"Hello human. I am mouse. I like to die."

"Noooo mouse don't die I put months worth of work into you."

"*liver failure*"

 

[Tutmos]

I sometimes think I'd rather work with yeast. Try figuring out if the LFP (local field potentials) in a primate prefrontal cortex recording mean anything relative to the trials and spike activity of the monkey. Oh teach yourself Matlab and isolate the LFP bands buried in the raw signal while you're at it. This may be no big deal for an EE with signal processing background, but I have none. It's really amazing stuff to work on and to have any part in trying to solve a fundamental process of how our brain works is beyond my expectations but at the same time it often feels like I'm trying to apply order to complete chaos, that in the end, may have no real purpose at all.

I've had two glasses of wine to put everything in appropriate perspective.

Edit: Oddly enough my interviewer never asked a word about my research work in this even though I'm now at 10 semester credits on the project.

 

[SteinUmStein]

oh, cell contamination - bacterial or fungus - is the absolute worst. it kills any collaborative environment of the lab and people start finger pointing. always takes up 30 minutes of a lab minute in a really serious mood. people stop sharing medias and you have to make up your own, which fills up the 4C really quickly. and then you run through sterile filters like whitney does pipes

This. 👎

So tired of bench work...

 

[SteinUmStein]

"Hello human. I am mouse. I like to die."

"Noooo mouse don't die I put months worth of work into you."

"*liver failure*"

Aww sorry, that really sucks. 🙁

 

[lorenzomicron]

I sometimes think I'd rather work with yeast. Try figuring out if the LFP (local field potentials) in a primate prefrontal cortex recording mean anything relative to the trials and spike activity of the monkey. Oh teach yourself Matlab and isolate the LFP bands buried in the raw signal while you're at it. This may be no big deal for an EE with signal processing background, but I have none. It's really amazing stuff to work on and to have any part in trying to solve a fundamental process of how our brain works is beyond my expectations but at the same time it often feels like I'm trying to apply order to complete chaos, that in the end, may have no real purpose at all.

I've had two glasses of wine to put everything in appropriate perspective.

Edit: Oddly enough my interviewer never asked a word about my research work in this even though I'm now at 10 semester credits on the project.

I recommend definitely not writing your own code. As a BME in Signal Processing, i think you should just consult with EE or use the toolboxes.

 

[Tutmos]

I recommend definitely not writing your own code. As a BME in Signal Processing, i think you should just consult with EE or use the toolboxes.

I'm using the Pwelch function in Matlab to act as a fast fourier transform or PSD (power spectral density). That seems to be working pretty well but I'm having problems right now with it shifting frequency scales based on the sample points (time period) I'm analyzing. I think I might have to downsample the longer time periods to present the same sample number of data points to maintain the scale, but that seems silly that pwelch can't auto scale the frequencies based on the number of sample points presented. It's likely user error on my part.

I could never, well I shouldn't say never, but not in the near future create my own functions to replicate what some of these canned functions do. It was bad enough trying to figure basic matlab out without the thought of having to create a function.

 

[Hifey]

And THIS is why I went into clinical research instead. 🙂

Seriously though to those of you who stick it out and publish...you have my respect!

 

[lorenzomicron]

I'm using the Pwelch function in Matlab to act as a fast fourier transform or PSD (power spectral density). That seems to be working pretty well but I'm having problems right now with it shifting frequency scales based on the sample points (time period) I'm analyzing. I think I might have to downsample the longer time periods to present the same sample number of data points to maintain the scale, but that seems silly that pwelch can't auto scale the frequencies based on the number of sample points presented. It's likely user error on my part.

I could never, well I shouldn't say never, but not in the near future create my own functions to replicate what some of these canned functions do. It was bad enough trying to figure basic matlab out without the thought of having to create a function.

If possible, not sure what kind of data sampling and structure you are dealing with, but the best option in all cases is to indeed downsample so that it's aligned. Or if you don't need to run the data congruently, just window the data (i.e. segment) so that the signal will run more smoothly.

I'm kinda curious about why your time scales are shifting? I've not used the pwelch algorithm, but is there a paritcular reason you're not just using the dft algorithm?

 

[Tutmos]

If possible, not sure what kind of data sampling and structure you are dealing with, but the best option in all cases is to indeed downsample so that it's aligned. Or if you don't need to run the data congruently, just window the data (i.e. segment) so that the signal will run more smoothly.

I'm kinda curious about why your time scales are shifting? I've not used the pwelch algorithm, but is there a particular reason you're not just using the dft algorithm?

Thanks for the conversation. It sparked what I think might be my solution. I'm looking at when certain frequencies are present so I do need all the scales to match up, including time. When you mentioned window it suddenly occurred to me that Pwelch automatically breaks the data up into 8 windows and stacks them on top of each other. If the window size isn't adjusted relative to sample size then each window will include x times more data, so this is likely what was skewing my window. Of course I better not get too confident that this is the solution or Murphy will bite me in the behind. The window assumes if I'm increasing data points that I'm just changing resolution, not length of time of the sample. Woohoo. Thanks for the idea.

 

[CapnCrunch]

My roomate told me that the ticks he was working with were not stored at the proper humidity (error on the company) so all the ticks they were working either went into their dormant phase or just died. So now the lab has to spend some $5,000+ and wait 4 weeks to buy and cultivate these damn insects. Luckily, I don't work with bugs. 😉

 

[RogueUnicorn]

If you have done research, you understand what it feels like to repeatedly bang your head against a wall to try to get some assay to work. For example:

Today I just found out that all of the cloning I've been attempting to do to get this protein expressed will not work because the initial DNA we had was bad. That's 4 weeks of my life I will not get back.

I'm sure other people had similar stories, and I would love to hear them. Feel free to vent about the horrors of bench research!

this EXACT thing happened to me, except it took 4 months to figure out.

 

[ShinyDome19]

Today I just found out that all of the cloning I've been attempting to do to get this protein expressed will not work because the initial DNA we had was bad. That's 4 weeks of my life I will not get back.

heh, 4 weeks isnt too bad. I was specifically hired at my current position for my experience with PCR and cloning. Well, the lab that hired me had an associated lab that had been working on a real-time PCR for two years and they couldnt get it to work. They tried for a year, then decided to simply send it out and get it done while they tried to get it to work in the lab...This cost them mega bucks every time. On my first day of work, I look at their protocol...the idiots (2 PhD's who primarily work on animals) forgot a little simple thing called a denaturing step in the PCR cycle ...worked like a charm after that.

 

[rockaction]

heh, 4 weeks isnt too bad. I was specifically hired at my current position for my experience with PCR and cloning. Well, the lab that hired me had an associated lab that had been working on a real-time PCR for two years and they couldnt get it to work. They tried for a year, then decided to simply send it out and get it done while they tried to get it to work in the lab...This cost them mega bucks every time. On my first day of work, I look at their protocol...the idiots (2 PhD's who primarily work on animals) forgot a little simple thing called a denaturing step in the PCR cycle ...worked like a charm after that.

😱 wow

 

[Tutmos]

heh, 4 weeks isnt too bad. I was specifically hired at my current position for my experience with PCR and cloning. Well, the lab that hired me had an associated lab that had been working on a real-time PCR for two years and they couldnt get it to work. They tried for a year, then decided to simply send it out and get it done while they tried to get it to work in the lab...This cost them mega bucks every time. On my first day of work, I look at their protocol...the idiots (2 PhD's who primarily work on animals) forgot a little simple thing called a denaturing step in the PCR cycle ...worked like a charm after that.

That's a riot! That's what you call getting the cart before the horse. For some reason I'm betting their appreciation won't be expressed in public. 😛

 

[ShinyDome19]

Well, in their defense, they did bench work secondary to their animal work..but, both were extremely grateful for my 5 minutes of help. One actually brought me a bunch of Chinese cookies his wife made for me...they were yummy.

 

[4X4MD]

my research currently consists of washing beakers, autoclaving, and making diluted solutions...oh joy (2 days in haha)

 

[ShinyDome19]

my research currently consists of washing beakers, autoclaving, and making diluted solutions...oh joy (2 days in haha)

nothing wrong with that...I did that my first year as an undergrad...I still wash dishes 7 years later. Gotta start some where.

 

[4X4MD]

nothing wrong with that...I did that my first year as an undergrad...I still wash dishes 7 years later. Gotta start some where.

yeah, i think my PI may be a little pissed at me...she took me on my freshman year through a special program reserved for the top 100 entering freshman across the entire campus and instead of spending time in the lab (paid time), i spent my time in bars...needless to say, i never put in any effort and quit the job really quickly because of the time demands...

im working on my redemption and im hoping my inquisitive nature will allow me to contribute to the research team. right now, tho, im partnered with a graduate student who would rather see me do the dishes than learn a single thing...but i pick things up pretty quickly

also, google is your best friend😀