A follow-up about Western Blots
A follow-up about Western Blots
Led by a comment from a reader about Jikkyleaks' newest incredible STINKY CHEESE leak
Jikkyleaks did some fabuloid work this week my friends. Here’s some background in the form of the original ‘tweet’.
I saw this problem in the original BREAKING CHEESE piece that Jikkyleaks posted and you can (and should) read about that here. Those Western Blot results looked like this:
I noted this on my own Twitter feed for posterity. It is quite amazing to me looking back how my language revealed something remarkable about the potential true story behind these ‘Western Blots’. I’ll get to that.
A Midwestern Doctor wrote this up in a Substack so well, that I have nothing much else to add. You can read that Substack here.
Notice I wrote ‘much else’. Of course I have an add-on!
So again, I hadn’t realized the deeper truth in my words that I posted in reference to the Westerns produced by Pfizer shown above. I wrote:
There is not a human alive who could produce a Western like this.
Human, eh. I noticed in the comments of A Midwestern Doctor’s Substack that someone wrote something about Western Blot automation that I was not aware of. They wrote:
Writes Overdue Revolution’s Newsletter
I sent this article to a friend who works in biotech, this is what he said:
"One of our instruments is an automated western blotting instrument. The data doesn’t come out in traditional form, it comes out as an electropherogram. And then we have software convert the signal peaks from the electropherogram into an artificial representation of a traditional blot. Pfizer has 20 of those instruments."
He linked this particular instrument as an example: https://www.bio-techne.com/pdf-download-arena-document/brochure/pl6-0002
Not to defend Pfizer, but simply the computer generated nature of the Western blots cannot be used to prove fraud.
Now this made me do a double-take because first of all, I didn’t know this was a thing. Apparently, at least in Pfizer land, it is. The reason I am kind of appalled by this automation of Western Blotting is because of the nature of a Western Blot. Let me explain. And by the way, not a word of a lie, the machine they use (according to the link in the comment above) is called Jess.
The reason you may have heard me say that I hate Western Blots is because they are extremely time-consuming. They involve multiple steps and lots of fun buffers and chemicals, like methanol, that you have to make up each time. Well I did because I like fresh buffers. So very many reagents to prepare, yes, and don’t forget the protein preparation that precedes the Western itself! If you’re lucky, you can buy nitrocellulose gels already snuggled into a pre-made case. But, then you have to crack that case open once the gel has ‘run’, and the suction created due to the nature of the gel makes it really hard to open that case and not rip the gel. It took me a long time to develop my special hands-on technique to crack that case without ever ripping my gel. Then came removing bubbles whilst stacking that Western sandwich (Homer moment: mmmm Western sandwich) with that gel all snuggly inside. Hashtag Western Blotters know.
Here’s a lovely British narrator describing the Western Blot protocol in case my journey above was too jibberishy. They do not using ice cold transfer buffer in this demo video but, I suppose I’ll let it pass. Anyway watch the video and think about how important the human is in this multi-step methodology.
It is almost funny to me that someone, somewhere, has automated - even the endpoint - of a Western Blot because, and call me old-fashioned, I will NEVER be convinced that any stage of the Western Blot can be replaced by a machine or non-human tool. It is from the moving, thinking being, with hands and a brain that the end point is reached, which is that little thin white membrane with those imperfect bands on it, that always seemed to me to be so anticlimactic after all that work and time. Nevertheless, standing in that room with the camera to take that photo of the membrane was exciting because, if you did a good job, and your proteins of interest were ‘revealed’, you’d see that pretty immediately.
I am sure that there isn’t a human alive who has done a lot of Westerns who hasn’t jokingly thought, ‘Gee, I wish I could just draw those lines where I think they should be. That would save me so much time!’ Or maybe that’s just me. And no, I never did that.
So to be clear, robo-Jess is new technology and a new machine to me, which is funny because she’s Jess. She functions to automate fluorescence and chemiluminscence detection, protein normalization and Western Blot imaging. So Jess can be used only at the endpoint of the Western Blot protocol - after all of the hard work - to translate the ‘lines’ on the membrane to an image.
Jess seems somewhat dreamy to me, on one hand. How gorgeous and perfect those results look! But here’s the thing, taking the human out of the equation makes me nervous. Here’s what auto-Jess can do:
Just load your samples and reagents into the microplate, and Jess does the rest. She separates your protein by size and precisely manages antibody additions, incubations, washes and even the detection steps. Come back to fully analyzed and quantitated results in just 3 hours.
Robo-Jess ‘eliminates many of the tedious, error-prone steps’ involved in the Western Blot protocol. Ok. But human Jess does the same by being careful and experienced. And who cares if it’s not perfect: that kind of makes the results more reliable, if you ask me. It’s like those rough edged, different sized cookies that grandma used to make, as opposed to those cookies you buy that were made in a factory - the factory cookies may all be the same size and have the same number of chocolate chips per cookie, but they are also, in many cases, full of chemicals and preservatives and taste… weird.
So even though robo-Jess seems like a dream, I am firm in my belief that human-Jess cannot be removed from the Western Blot equation, speaking for myself as human-Jess. There are probably only a few people left who will agree with me on this, but I stand firm. There are just some things that you can’t - and dare I say, shouldn’t - automate. Cookie production is one, and Western Blots are another. Sorry Pfizer. Even with all of your money and robo-Jess, I personally, as a biochemist and protein biologist, will always prefer the roughness and authenticity of the hand-made Western Blot.
I suppose I should end this Substack by saying that although the perfect, gorgeous Westerns produced by Pfizer may not be fake, per se, they certainly are not ‘reliable’ either - by my assessment. Perhaps preferable is a less strong word to use, but still, for all of the reasons I have given here, their results are just imposters and must be redone by a human. Again, many will disagree with me on this, but as a hardcore, old-school science nerd, technique and hands-on approaches cannot be replaced.
I don’t want a machine making my cookies, and I don’t want a machine doing my Westerns. Even if she does have a cool name.
Robo-Jess fills in our western blots, like how Dominion fills in our voting blots.
Jess (the real one not the robo-one) - thank you for summing this up so perfectly! The issue for me is that these AWBs (which are not Western blots but are essentially interpreted electropherograms displayed in a nice even plot) are just so easy to falsify. You can't use these as proof of anything. They are internal screening devices to look for stuff which, when you publish, you'd better have the real Western otherwise your reviewers are (should be) asking questions.
Which is exactly what the FDA didn't do.