13 replies to this topic

[BMF]

I have a paper reading course in McGill University for PhD students.

A Professor is in charge of this course. I usualy find a msitake in primers in paper in average one out of 15 papers, so that these primers do not bind to that gene of interest in paper and so that data is fake.

 

But this professor laughed at me and siad in front of students:  Haha ha  he (me) likes primers.

Actualy this is not the first one. Also I found that he does not know what primer is and what it is for. and how to check primer.

 

What should I do? After 30 years from invention of primer?

Edited by memari, 29 April 2015 - 08:11 AM.

[r.rosati]

First of all, I think it's awesome that you check the primers for consistency. Well done.

Now, about your situation, first be absolutely sure that the primers really won't bind to the target. Sometimes, for various reasons, the designed primers have a few mismatches. In this case the primer would bind anyways (and sometimes with better performance, or with specificity for an allele...). These mismatched primers might, perhaps, not show up in a default BLAST search (depending on your parameters) - they will surely not be found by a literal search. Also, sometimes primers were designed from a sequence that wasn't perfect and that doesn't match the official human genome - this isn't so rare in old papers.

So I would first check the literature (i.e. Google Scholar) for other occurrences of the same primer sequence (not a guarantee of correctness, but it's a good sign); then check whether the primer really binds somewhere else, or whether it has a few mismatches; and if it's mismatched, consider if it could extend anyways. If the differences are in favor of it not binding, then yes it might well be a typo... sometimes it happens. It doesn't have to be necessarily fake data.

Edited by r.rosati, 29 April 2015 - 10:40 AM.

[BMF]

when you do Primer-Blast for a primer from a 13 impact factor journal in 2014.

And these primers bind everything except the gene of interest, you say again this:

 

"sometimes it happens. It doesn't have to be necessarily fake data.."

[Trof]

I'm not sure if you are supposed to find fake papers or what, but true is many times there, the primers are wrong or have a mismatch not mentioned (mismatch can be tolerated in old papers, but not in new).

 

I think truth is, people just plainly don't care much about correct sequence and use primers that "their grandma used " (haha) and never check them later cause they "work", or they mistakenly switch the sequences for other primers, or for a reason of keeping them secret they intentionally don't publish the right ones, because this is so common that you can always say it was a mistake.

 

This of course is not true in "methodic" paper, since everyone will try it and find it doesn't work.. but in common papers.. primes, who cares..

 

My boss says, the better impact the less they care about what details of PCR you state, if you are putting it into lesser impact they will want to know exactly the sequences and conditions, if you were to publish in Nature? No one cares..

So mistakes are common and even if it's sometimes intentional probably, it doesn't mean the data are fake. But it is wrong to do so. Without primer sequence you cannot check for potential problems thay the study may not have addressed (like... later will be found, that these primers bind nonspecifically... it is important to include this right).
But I suppose it is considered a minor problem, faka data on the other hand is a serious scientific misconduct.

[pito]

do you have some examples of papers with mistakes? I am curious to see.

But yes: it happens.

[BMF]

Ok this is a sample from Nature:

 

http://www.nature.co...ary-information

Primers are in this link:

Supplementary Table 10. Primer sequences

http://www.nature.co...ure13684-s2.zip

 

Just check the B2m for mice and CXCL2.

B2m does not bind at all.

 

https://i.imgur.com/bDsRUJu.jpg

https://i.imgur.com/t9qGkHl.jpg

https://i.imgur.com/Mu3qRtc.jpg

 

 

Babak

Edited by memari, 04 May 2015 - 09:32 AM.

[r.rosati]

The reverse primer AGCTATCTAGGATATTTCCAATTTTTGAA indeed lacks a T to match the official sequence (which is AGCTATCTAGGATATTTCCAATTTTTTGAA).

Still, it matches the alternate Celera assembly. And I think it would extend. I can see a couple of other papers that used the same reverse primer (and curiously, no papers that used the "reference" sequence). I think they just bought the same primer that they saw published, and I can see how it can extend after all.

 

Check: try BLASTing this,

ATGCACGCAGAAAGAAATAGCAA

NNNNNNNNNNNNNNNNNNNNNNNNNNNNNN

AGCTATCTAGGATATTTCCAATTTTTGAA

 

Algorithm: blastn, expect: 1000, word size: 7.

 

 

I find:

http://blast.ncbi.nl...RID=MFND3WSB01R

You can see a good match for the Celera genome and a partial match for the refseq genome, but extend it a bit and you'll see that the problem is one T.

Edited by r.rosati, 04 May 2015 - 12:41 PM.

[pito]

The reverse primer AGCTATCTAGGATATTTCCAATTTTTGAA indeed lacks a T to match the official sequence (which is AGCTATCTAGGATATTTCCAATTTTTTGAA).

Still, it matches the alternate Celera assembly. And I think it would extend. I can see a couple of other papers that used the same reverse primer (and curiously, no papers that used the "reference" sequence). I think they just bought the same primer that they saw published, and I can see how it can extend after all.

 

Check: try BLASTing this,

ATGCACGCAGAAAGAAATAGCAA

NNNNNNNNNNNNNNNNNNNNNNNNNNNNNN

AGCTATCTAGGATATTTCCAATTTTTGAA

 

Algorithm: blastn, expect: 1000, word size: 7.

 

 

I find:

http://blast.ncbi.nl...RID=MFND3WSB01R

You can see a good match for the Celera genome and a partial match for the refseq genome, but extend it a bit and you'll see that the problem is one T.

 

Indeed, got the same blasting the reverse primer: https://blast.ncbi.n...lnHdr_144227219

 

However I find it weird they do not refer to what sequence they used.

If you check one of the sequences/references they make, you end up with another sequence than this one.

 

They should be better in mentioning what exactly they used. It is sometimes frustrating to figure these things out yourself.

[BMF]

Another funny papers that all primers are wrong except one:

 

Impact factor 6 and Cited by 188 since 2009 until now:

 

http://atvb.ahajourn...t/29/6/936.full

 

this paper should have been in retractionwatch.com

Edited by BMF, 11 February 2016 - 05:40 PM.

[pito]

Pretty funny indeed.

Only 1 primer (the reverse, second primer in the supplement) is indeed correct.

 

Another funny papers that all primers are wrong except one:

 

Impact factor 6 and Cited by 188 since 2009 until now:

 

http://atvb.ahajourn...t/29/6/936.full

 

this paper should have been in retractionwatch.com

[Trof]

I learned about these mistakes and "mistakes" as one of the very first thing at all in my scientific life.

 

On a masters program (that is basically the only college one there was at a time, Bc. programs were very rare on our universities then, you just got all 5 years at once and then masters), my first individual task on the diploma project. My supervisor gave me a paper about qPCR for a virus and asked me, what would I do first, when I want to use them. I finally guessed I should probably check them on some sample even if they are published to work and he said, "check them before ordering.. do you know what tools to use?" He told me and after some hours and rechecks I get back to him and I told him, surprised, that the published primers have a single bp mismatch that only aligns with one of the two virus strains. I was so shocked that they didn't mentioned it in the paper. My supervisor told me to design my own, correct then. Those were the first primers I ever designed

[Michael Starr]

I'm not sure if you are supposed to find fake papers or what, but true is many times there, the primers are wrong or have a mismatch not mentioned (mismatch can be tolerated in old papers, but not in new).

 

I think truth is, people just plainly don't care much about correct sequence and use primers that "their grandma used " (haha) and never check them later cause they "work", or they mistakenly switch the sequences for other primers, or for a reason of keeping them secret they intentionally don't publish the right ones, because this is so common that you can always say it was a mistake.

 

This of course is not true in "methodic" paper, since everyone will try it and find it doesn't work.. but in common papers.. primes, who cares..

 

My boss says, the better impact the less they care about what details of PCR you state, if you are putting it into lesser impact they will want to know exactly the sequences and conditions, if you were to publish in Nature? No one cares..

So mistakes are common and even if it's sometimes intentional probably, it doesn't mean the data are fake. But it is wrong to do so. Without primer sequence you cannot check for potential problems thay the study may not have addressed (like... later will be found, that these primers bind nonspecifically... it is important to include this right).
But I suppose it is considered a minor problem, faka data on the other hand is a serious scientific misconduct.

 

That is interesting, Trof. I would've thought the higher impact the journal, the MORE scrutiny you'd get about details like primers, since the readership is higher.

[Trof]

That is interesting, Trof. I would've thought the higher impact the journal, the MORE scrutiny you'd get about details like primers, since the readership is higher.

Yes and no. The high tier journals you can't just publish in like in any others. You need to have a name (or know someone to have a name or optionaly promise something so cool that the editors will be eager to push it). So, while you publish there, you are a well known leader in the filed, or rising star (or the paper has both) so noone bothers that much about the little details. They go for the cool story, many higher IF journals will reject saying the topic is not interesting enough or doesn't have a novel clear messaage, in other words, not cool. It's not that the data are bad, they just not that sexy to be in their journal.

I repetedly hate people publishing in Nature for using non-HGVS nomenclature for mutations for example. If you want to publish a devilishly difficult gene insertion in Human Mutation, you need to spend week in email correspondence with one of the curators to get even the notation right, because it is such a complicated mess. The journal would refuse to publish without a correct mutation name.
Now Nature? Who cares, it's sexy! (it is, though)

Then I sit with clinical diagnosticians and they ask me, which one is right.. the one by HGVS or the one everyone is using, even that famous paper in Nature.. what can you say?

You would think that top journals need to have a more rigorous control of the data, but they likely more rely on the people-basis than the data itself and more important is how cool the story is. Mostly they are right and the data are correct and cool at the same time. But sometimes, as with the case of the famous arsenic eating bacteria in Science.. it was soo cool and so anticipated, that it blinded them so much to miss the major flaws in the paper, and many objections were found within the second day the paper was out. Those would be caught by some sceptical-style reiewer, but for some reason they were not. Coolness won.

[BMF]

In Canada, educated people have low knowledge, especially authorities.

And the problem is they do not care too.

 

This is why the Canadian are successful in two jobs in US, because these jobs need low knowledge and talkative persons:

1) Comedian 2) Singer

 

Those who think Canada is the best and the Canada Health checks every thing, should see this:

 

Toxic jewelry: What’s in the cheap jewelry you buy (some had 100% cadmium even in Sears)

http://www.cbc.ca/ma...6/toxic-jewelry

 

 

Canada does not do anything. They just sit and wait for US authority's decisions.

These two happened during past 3 years:

If US warns for a contaminated beef, 3 days after that, Canada warns for it.

If US warns for Tsunami, 3 days after, Canada warns for it.

Edited by BMF, 30 March 2016 - 07:28 AM.