#SSHOw No. 2- GMO Maze
We hopefully guided you through the maze of the GMO Corn hysteria in the media. Here’s a link to some of the statistical issues. http://goo.gl/epcnr Next weeks #SSHOw will be on the mummy bird that I posted a while ago (http://goo.gl/sbzJq).
#ScienceSunday
Bad science → bad headlines
Today’s #SSHOw (http://goo.gl/0eGrh) will be discussing GMO corn and in particular a poorly done experiment/publication that sparked the media storm (Séralini et al Food and Toxicology 2012). http://goo.gl/5GOWa
Orac dissects the paper quite nicely, although I think he repeatedly says mice when he means rats. http://goo.gl/SSE2F
The two areas I will comment on are the tumor rat model and a statistical issue dealing with multiple comparisons.
Spontaneous Tumors in Rats
Orac points to a study in 1979 where 81% of the Sprague-Dawley rats, the same strain used in the Séralini paper, develop tumors. When you use an animal model looking at tumor development, you need to know the prevalence of spontaneous tumor development. The control group(s) have to be designed such that you can differentiate between “normal” spontaneous tumor development in the control groups vs. the experimental groups. Part of that design is having sufficient number of animals to have statistical power. Using previously published data, the authors could have done a power analysis to determine the proper sample size. For these types of studies, where you are not doing intricate daily or weekly interventions/experiments, i.e., just keeping the animals for long periods while looking for mortality, it is not uncommon to have 3-5 times the number of animals used in the Séralini study.
As an example, I had the privilege to collaborate with Prof Morris Pollard at Notre Dame who developed the Lobund-Wistar rat model. Lobund-Wistar rats spontaneously develop prostate adenocarcinoma (PA) at a mean age of 26 months. In the publication of the model, out of 72 L-W rats, 19 (26%) developed large PAs. Imagine if Prof Pollard only used 10 male rats as in the Séralini study.
The Lobund-Wistar rat model of prostate cancer.
Pollard M.
J Cell Biochem Suppl. 1992;16H:84-8.
http://www.ncbi.nlm.nih.gov/pubmed/1289678
We use a transgenic mouse model for spontaneous ductal carcinoma in situ. Invasive carcinomas develop in 100% of the mice. The mice are called SV40-Tag mice based on the C3(1)/Tag mice. SV40-Tag stands for Simian virus 40 T-antigen, a trans-activating protein, which are essential for viral gene expression.
Green et al Oncogene 2000
http://www.ncbi.nlm.nih.gov/pubmed/10713685
The point is you have to know the tumor prevalence in the rodent model you are using and plan the control groups accordingly.
Multiple Comparisons/Sample Size
The study mentions that they use Discriminant Analysis (DA) to partition groups, i.e., you lump all the variables (factors) together and use DA to flesh out which factors influence the outcome, e.g., tumor size, biochemical markers, etc. In image analysis we use Linear Discriminant Analysis (LDA,http://en.wikipedia.org/wiki/Discriminant_analysis, http://goo.gl/oyNzh) to segment (classify) pixels. Say you want to automatically segment tumor from normal tissue using several image types of the same sample. You have thousands of pixels to work with, not 10. The method isn’t robust with 10 or less samples in the 20 groups used (note I separated the male and female groups in the Séralini study). Also, in the context of machine learning, you have to have a training set. In my example you give the program a set of pixels that you know belong to each group before testing the pixels you want to classify.
A quick review on null hypothesis testing. A type I error is when the null hypothesis is true but is rejected, i.e., a false positive. A type II error is when the null hypothesis is false but is incorrectly accepted as true, a false negative. Remember the null hypothesis can never be proven.
Here are examples from the Wiki:
Suppose the treatment is a new way of teaching writing to students, and the control is the standard way of teaching writing. Students in the two groups can be compared in terms of grammar, spelling, organization, content, and so on. As more attributes are compared, it becomes more likely that the treatment and control groups will appear to differ on at least one attribute by random chance alone.
Suppose we consider the efficacy of a drug in terms of the reduction of any one of a number of disease symptoms. As more symptoms are considered, it becomes more likely that the drug will appear to be an improvement over existing drugs in terms of at least one symptom.
Suppose we consider the safety of a drug in terms of the occurrences of different types of side effects. As more types of side effects are considered, it becomes more likely that the new drug will appear to be less safe than existing drugs in terms of at least one side effect.
http://en.wikipedia.org/wiki/Multiple_comparisons
Statistical power is the probability of committing a type II error, false negative. Prof Pollard’s study has a statistical power of around 97% while the Séralini study is probably closer to 45%.
In Memoriam
I the process of digging up the study by Prof Pollard, I realized he had passed away. I met him when he was in his late 80s, to do an experiment for him at the University of Chicago. He was an impressive man and scientist. It is really a shame he is most often known for his son.
Pollard worked at all of these things until his very last days. “I can’t imagine doing anything else,” he said recently. “I think if you are doing something meaningful and important and you stop doing it, you’ll always look back with regret.”
L-W Rat pic: http://goo.gl/hzP6P
For #ScienceSunday curated by Allison Sekuler Rajini Rao Robby Bowles and me.
Two Towers, Two IP Stories
Here’s my Soapbox Science blog series post on nature.com blogs
It’s related to my Bench to bedside post. http://goo.gl/8xtDH I share two examples about intellectual property (IP) in an academic setting; one good and one bad.
Thanks Laura Wheeler for the opportunity to share my story and thoughts. Thanks Jerry Nguyen for introducing me to the Soapbox series via the #PhDelta posts.
http://blogs.nature.com/soapboxscience/2012/09/26/two-towers-two-ip-stories
Overselling preclinical results – Orac
Here’s an article by an anti-quack blogger extraordinaire. http://goo.gl/DRpQk It says a lot of what I said in my comments about the Telegraph article discussed here and the linked discussion within:
Bench to Bedside
Of course Orac does a much better job of taking the article apart.
Attached image:
A scanning electrom microscopic image of HIV. The glycoprotein complex on its surface enables the virus to attach to and fuse with target cells to initiate the infectious cycle.
#ScienceEveryday when it isn’t #ScienceSunday
Analysis of Meta-analysis
You’ve probably seen quite a few posts regarding the Stanford study that concluded that there was little advantage to consuming organic foods.
Annals of Int Med http://goo.gl/fv4fG
Gaythia Weis asked me to comment on her post about ta blog post about he study. http://goo.gl/rerki
Sorry to disappoint you all, but I haven’t had enough time to dig into this framing debate. It really isn’t my area anyway. However, I use statistics all of the time in my research. So I’m going to introduce what is meta-analysis and hopefully return to the articles later.
Meta-analysis is a method of contrasting and combining results from a large group of studies. That sounds like an advantage and it is, i.e., strength in numbers. However, there are disadvantages and meta-analyses can be done wrong. The biggest disadvantage is that each study has a purpose and study design that is most likely not identical. So the key is how do you decide which studies to include and which studies to exclude. In today’s #ScienceSunday SSHOw, I hope to discuss this in a little bit more detail. The Wiki is actually quite good.
http://en.wikipedia.org/wiki/Meta-analysis
Getting back to the pesticide story here’s the abstract from the article.
Background: The health benefits of organic foods are unclear.
Purpose: To review evidence comparing the health effects of organic and conventional foods.
Data Sources: MEDLINE (January 1966 to May 2011), EMBASE, CAB Direct, Agricola, TOXNET, Cochrane Library (January 1966 to May 2009), and bibliographies of retrieved articles.
Study Selection: English-language reports of comparisons of organically and conventionally grown food or of populations consuming these foods.
Data Extraction: 2 independent investigators extracted data on methods, health outcomes, and nutrient and contaminant levels.
Data Synthesis: 17 studies in humans and 223 studies of nutrient and contaminant levels in foods met inclusion criteria. Only 3 of the human studies examined clinical outcomes, finding no significant differences between populations by food type for allergic outcomes (eczema, wheeze, atopic sensitization) or symptomatic Campylobacter infection. Two studies reported significantly lower urinary pesticide levels among children consuming organic versus conventional diets, but studies of biomarker and nutrient levels in serum, urine, breast milk, and semen in adults did not identify clinically meaningful differences. All estimates of differences in nutrient and contaminant levels in foods were highly heterogeneous except for the estimate for phosphorus; phosphorus levels were significantly higher than in conventional produce, although this difference is not clinically significant. The risk for contamination with detectable pesticide residues was lower among organic than conventional produce (risk difference, 30% [CI, -37% to -23%]), but differences in risk for exceeding maximum allowed limits were small. Escherichia coli contamination risk did not differ between organic and conventional produce. Bacterial contamination of retail chicken and pork was common but unrelated to farming method. However, the risk for isolating bacteria resistant to 3 or more antibiotics was higher in conventional than in organic chicken and pork (risk difference, 33% [CI, 21% to 45%]).
Limitation: Studies were heterogeneous and limited in number, and publication bias may be present.
Conclusion: The published literature lacks strong evidence that organic foods are significantly more nutritious than conventional foods. Consumption of organic foods may reduce exposure to pesticide residues and antibiotic-resistant bacteria.
I don’t really have a problem with the description of their methods. Again, I hope to return to this and talk in more detail.
Pic Source: http://goo.gl/LQRzE
Deadly but wonderful
How’s that possible? These are great pics; almost art, except they are deadly diseases.
Originally shared by Gail Barnes
Stunning But Deadly
The transmission electron micrograph of the Ebola virus would not look out of place on the wall of an art gallery, while Smallpox resembles sushi. Fantastic electron micrographs allow one to get up close and personal with some of the world’s deadliest diseases. #sciencesunday
Open the Spigot for Science
#ScienceSunday would love to ensnare you in our web. I saw this electron microscope pic shared from Linda Hedrick and I remembered that there has been a lot of research to mimic spider web silk. The pictures below are the spigots of the spider where the silk is made/dispensed for making a web. The first is a false-color version from Visuals Unlimited (http://goo.gl/2Qgui).
From MicroAngela
These fingerlike spinnerets on spiders’ posterior abdomens (rear-ends) are used to extrude web silk. This silk is used to weave webs, snares, shelters, and/or egg sacs. Each species of spider has a distinctive web form. Spider silk is a fibrous protein that is secreted as a fluid and which, when stretched, forms a polymer that is stronger than steel! A spider can spin more than one kind of silk to customize its web. For example, the spider makes some parts of its web not sticky so that it can run across it and not get caught! Some spiders do not weave webs at all, but actively hunt for food.
A tidy, clean web indicates a spider is present. A dusty web (cobweb) usually means it is old and unused. Spiders are useful in keeping down pest insect populations. Have you made friends with a spider, lately?
This picture was taken by a fourth grade class visiting my lab. It is magnified about 1,500 times
http://www5.pbrc.hawaii.edu/microangela/spigot.htm
The same image can be found at How Stuff Works, with ton of info: http://goo.gl/d0Pr0
Some folks at Rochester detailed their SEM study on spider silk on this webpage:
Here’s a clever use of spider silk. Dip the silk in organic silicate, then bake it so that the silk burns away leaving a small, hollow fiber optic wire.
Finally, here’s artificial spider silk from bacteria
I leave you with No Doubt
You think that we connect
That the chemistry’s correct
Your words walk right through my ears
Presuming I like what I hear
And now I’m stuck in the web
You’re spinning
You’ve got me for your prey
ScienceSunday curated by Allison Sekuler Rajini Rao Robby Bowles and me.
Edit also for #SpiderSunday curated by Kimberly Hosey and Chris Mallory
Bench to Bedside
A little background on translational research or medicine, often also called “bench to bedside”. In a post by Max Huijgen (http://goo.gl/VNT3Z) there was a lively discussion about an article in the Telegraph about a seemingly miracle cancer drug sitting in a freezer in Sweden. The article stated that because the scientist had published the work, no pharmaceutical company would touch it, to bring it to market. The idea was, no patent means no profit. A large part of the discussion was about crowdfunding for this frozen drug.
There were three things that Rajini Rao some others, and I tried to explain, about the article and the conversation. First, the cost of getting a drug “from bench to bedside” was grossly underestimated by people suggesting crowdfunding. Here’s the reference Rajini provided regarding the cost of clinical trials (http://goo.gl/YXD55). It can be $28k per patient, just to get started, i.e., Phase I. Which leads to the second issue; lack of understanding what is drug development or discovery. How do you get a compound from idea, to bench work, and finally to market. The last issue is the treatment in the story itself.
The figures below are from an excellent overview of the process of drug development. The example is for cancer research but it is applicable to essentially all drug discovery. So when you hear scientist like me, talk about Phase III clinical trials or INDs, you’ll know what they mean. http://goo.gl/ZF88v
There is also a good, albeit a bit old, review on translation research.
Review: Translational science: past, present, and future by S. H. Curry
BioTechniques 44, 2008 http://goo.gl/B5DV6
In that review, they mention acetaminophen being published and therefore the university couldn’t capitalize on the discovery. If the premise of the Telegraph article were true then acetaminophen should not be in every medicine cabinet today. Clearly the inability to directly patent a compound does not preclude getting it to market. I say directly because acetaminophen was patented at some point.
History of acetaminophen or paracetamol
http://en.wikipedia.org/wiki/Paracetamol#History
Getting back to the Telegraph article and the miracle drug. The Swedish group’s current paper, that has in vivo data (the rest are in vitro) has a 40% survival rate at 100 days, using six mice per group. I also find it curious that the virus is supposed to be targeted but they did not inject it systemically. It was injected directly into the tumor.
J Virol. 2011 Dec;85(24):13114-23.
http://www.ncbi.nlm.nih.gov/pubmed/21957304
Although it is interesting and promising work, I wouldn’t put any money towards it with such a small study. Typically, much more pre-clinical work is done before even thinking of putting a therapeutic agent into humans. As others stated in the thread, there are tons of therapies that cure mice but do not work in humans.
The adenovirus that I have worked on is not targeted per se. It has to be injected directly into the tumor. However, the gene is activated only where you irradiate. The tumor gets a double whamy: radiation and damage from the gene therapy. It was in a phase 1 trial, published here: Clin Cancer Res. 2004 Sep 1;10(17):5747-53. http://www.ncbi.nlm.nih.gov/pubmed/15355902?
I want to emphasize that I have nothing against the research done by the Swedish group. I do have an issue with the Telegraph article. The author talks about his keen investigative skills to find the Swedish group but yet doesn’t do due diligence in finding out the true potential of this therapy. Again, it’s interesting work but a 40% survival rate in mice, in a small single study, is not a miracle drug. The author even mentions the Amgen “investigation” (but gives no reference) as a reason to be cautious. It’s actually a commentary, not a peer-reviewed publication. My commentary about it was posted a while ago, here: http://goo.gl/98HmX? One should be cautious because a lot promising pre-clinical studies don’t pan out, not because of a commentary in reproducibility. Reproducibility is certainly important and is a problem in some areas of research. It’s frustrating that the post follows the same tone as the Telegraph article. However, as a proponent of science, I’m happy that there was a lively discussion and there was an opportunity to educate.
#ScienceEveryday when it isn’t #ScienceSunday
