Editorializing on Open Access

I've agreed to write a 250-word contribution about open access scientific publishing, for our Faculty of Science magazine.  I think the main readership is science alumni.  Below are some stream-of-consciousness ideas I might incorporate:

• The internet is opening up science to public view.
•  Publishing and critical review by other experts is the core of science.  Nothing we do in the lab matters until it's publicly available for evaluation.
• Papers are published in journals, where they must first pass evaluation by peer reviewers.  
• One function of journals has been to carry out peer review, and the other to physically publish and distribute the approved papers by mailing out issues of the journal.  
• We don't need to change peer review, just the publication process.
• Open access has nothing to do with peer review.
• Access to these issues was by subscription, with research libraries paying very high rates (discount subscriptions, still very expensive, are available but typically only held by researchers working in that field).
• Before the internet, each additional reader added a direct cost, as papers had to be physically printed and distributed.  With online-only publication, the costs are largely independent of readership.
• The taxpayer pays for most scientific research, and they should be able to see what we've done with their money.
• Journalists need open access to research papers.
• The subscription-based publication system scientists are used to is itself relatively new.
• Publishing research for profit probably won't go away completely, because both publishers and scientists have vested interests (profits and prestige respectively) that are independent of the intellectual purpose of scientific publication. 

The above is just a partial list of factoids, and it's already almost 250 words!   I need to cut it down and find a theme that makes it interesting.  Would my point be that the taxpayers are being bilked?  That scientific progress is being obstructed?  That old-media scientific publishing is a scam and a scandal?  Should I pick a specific journal as an example of how bad things are?  Hmmm, something published by Elsevier?

Examples of pay-per-view access:

• Science wants $15 for 24 hours of access to a 1992 article on the antibiotic resistance crisis.
• The Lancet wants $31.50 for 24 hours of access to a 2001 article on antibiotic resistance in biofilms.
• The journal of the American Medical Association wants $30
• J Bact (ASM) wants $20, but only for articles less than a year old.
• Journal of Clinical Oncology wants $22 for access to an article on chemotherapy resistance.
• Cancer Treatment Reviews (Elsevier) has no purchase option for single articles.  Nor does Gastroenterology.  Nor Diabetes Research. Nor Neurobiology of Aging.
• Insect Biochemistry and Molecular Biology $41.95
• Obesity $32
• Physics Letters A:  $31.50
• Physics Review Letters:  $25
• Wiley won't even tell you the price until you register with them.
• Journal of Electron Microscopy:  (OUP) $42.48
• Journal of Nanoscience and Nanotechnology $113  (for 7 pages)
• Nature $18;  Nature Physics:  $32.
• http://content.karger.com/ProdukteCytogenetic and Genome Research (Karger) $38, or $26.50 with an 'Pay-per-View account (for a paper from 1992!).

Got it!  "Pay-per-view isn't just for Porn"


Oops. UBC isn't very keen on having 'porn' in the title, so I'll change it to 'sports' and mention porn in the text.

Annual Reviews pay-per-view is inconsistent.  Many articles appear to be freely available, but a 1997 one from Ann Rev Medicine costs $20 for 24-hr access, or 99¢ from a site I'd never heard of called 'DeepDyve.  ARM says DeepDyve is "the largest online rental service for scientific, technical, medical, and scholarly research articles."  I just checked it out; it deserves its own post.

BioEssays gets it!

The Editor of BioEssays, Andrew Moore, has just posted several long comments (essay-length in total) about BioEssay's policies concerning blogs and other informal scientific communications.  You'll find them at Zen Faulkes' excellent Neurodojo blog.

Dr. Moore's comments are very reassuring; BioEssays clearly recognizes the important contributions that post-publication peer review can make to peer-reviewed science.

Even the New York Times is fearmongering about radioactivity from Japan

The New Your Times has an article today headed "Radiation Plume Reaches U.S., but Is Said to Pose no Risk"  The article itself goes on to explain that all the experts and all the measuring equipment agree that the amount of radiation in the plume is miniscule, detectable only because the instruments are extremely sensitive, and that it poses no health risk.

Why then include the words 'is said to' in the headline?  They dramatically weaken the message, implying that, although someone says that the plume poses no risk, we don't know whether or not it actually poses any risk.

I realize that the article's author, William Broad, probably had no say in the headline.  But the Times editor responsible should be ashamed of misleading the public in this way.

Response from Drs. McDermott and Rosen about their arsenic paper

The authors of the arsenic review paper I described in the previous post have put up a response to my comment.  It's a bit hard to find the comments for this paper, so here's a link.  

Basically, they don't read online science that hasn't been peer-reviewed, and thus the content of their article was written independently of the post-publication discussion of the Wolfe-Simon paper.  They would not have cited this discussion in any case because it was neither peer-reviewed nor personally communicated to them.  They feel that although some online communication about science may be of high quality, its value is negated by the simultaneous presence of low-quality material.

Comment posted on Rosen paper

Last week a paper appeared on the BioEssays Early View page, titled Life and death with arsenic, and subtitled Arsenic life: An analysis of the recent report "A bacterium that can grow by using arsenic instead of phosphorus"*.  It's a good overview of the properties of arsenic that are relevant to biology, and of many of the flaws in the Wolfe-Simon paper.

I've just posted the following comment on the BioEssays Discussions page for this paper:

The authors should not have dismissed the  extensive and well-supported online evaluation of the Wolfe-Simon paper with the single sentence "This paper has generated significant commentary, often as anonymous electronic communications."

Many of the points that Rosen et al. present without attribution were first raised by other researchers in  online articles, in authoritative blogs and in signed comments.  The subtitle "Arsenic life" is even taken from the discussion's Twitter hashtag "#arseniclife".

It hasn't appeared yet - comments have to be first approved by the editors.

* This link just takes you to the Abstract.  I can't tell whether the paper is open-access or not - the authors would have to have paid $3000 for this, so probably not.  Sorry.

IHMC11 last session, Friday afternoon

Bin Hu (Beijing Genomics Institute): A new way to research into human gut microbiome.
Unfortunately he's so far only giving an overview of the general microbiome approach, and the history of less-efficient methods (cloning...).  HMP (6 body sites) and MetaHIT (Human Intestinal Tract) projects...

The BGI has 137 Illumina HiSeq sequencers and 27 SOLID 4 systems!  They have data for 425 people, investigating food, obesity, IBD, diabetes, HBV, other viruses.  But "We need more data."

(I had to move farther back to find a free place to plug my laptop in, and now I can't read the smallish print on his slides.)

Announces The International Earth Microbiome project!


Bruce Birren: "High-throughput 16S sequencing for human microgenomics"  But this talk seems to be about quality control - very interesting.  Generating the reference genome set for the HMP.

Clinical team established protocols to screen patients for 'normality' and for collecting biological material from them.  >17,000 primary specimens.  Converted these into DNA for the sequencing centers.

Consistency of results:  Controls: same samples to all 4 centers.  The center-to-center variation swamped the biological variation!  So needed to standardize the biological and informatics protocols.  The final protocols may not be the BEST, but they're the most reproducible.

Accuracy of results (this is 16S results from 454):  21 organisms, range of %G+C.  Yet another heat map.  Gave centers a mock-community.  Good correlation, but 5% of sequences weren't from the input organisms.  Problem turned out to be 454 chimeric reads.   Previously undetected because algorithms weren't sensitive enough. So developed a new tool: ChimeraSlayer.

Problem with classifying reads form different biological sites.  Not fixed yet but will be.  Problems with overestimating diversity:  13 input species gave >200 apparent species (OTUs)!

Data:  first 24 people:  Can we define a 'core' microbiome?  Plot percent of samples the organism is in, vs abundance of each organism.  The commonly-present organisms are not necessarily the common organisms.  How distinct are the microbiomes of the different body sites.  Many genera found in only one site (= specialist genera), but many present in all sites.  Specialist genera are not very abundant.

Overall, can't readily connect the 16S data to the biology.  Need not just more metasequencing but more sequences of reference genomes.

Karen Nelson (JCVI):  Tools for the trade.  Changed her talk to highlight gaps in info in this meeting.

Working groups and body-site specific working groups have been meeting every week for past 3 years, to generate list of organisms needed for scaffolding this kind of data, to serve as a community resource.

Original goal of 1000 reference genomes now 3000, also viruses and phage and protists.  Tree of 1400 species in the pipeline for genome sequencing.

NSF-funded study on role of diet and gut microbiota in non-human primates.

Claire Fraser-Liggett:  The role of gut microbiota in obesity in the Amish.  Metabolic syndrome an issue.  In general, the published papers haven't reached consistent conclusions.

Old-order Amish are a closed inbred community with well-documented ancestry (large families, 14 generations).  Diet quite consistent across families, not much snacking, not much prescription drug use.  GWAS data available for nearly 4000 people.

Lean, overweight, obese and metabolic-syndrome obese people, two samples > 3 months apart.  At level of bacterial family, can't tell the four groups apart.  Not at level of phylogenetic abundance.  Bacteroides:Firmicutes ratios differ but not significantly.  Very stable over time, maybe due to homogeneous Amish lifestyle.  No significant differences in diversity.  Household and heritability effects?  Not for household or siblings.

Core microbiome: maybe 13 genera., Bacteroides and Firmicutes???  A couple of species significant:  Ruminococcus, Faecalibacterium (marker of inflammation?)  11 genera linked to cholesterol and serum triglyceride levels and BMI.  See correlations between specific bacteria and the GWAS studies on the people carrying them.

So far they've only done 16S  RNA studies - showing who is there.  Next is to find out what they are doing.  Conclusion might be that gut microbiome is irrelevant to obesity (at least in OOA), but maybe the superficial similarity will turn out to be quite different when metagenome and transcriptome analysis is done.


George Weinstock:  What's next for the human microbiome (next meeting in Paris)?  He thinks the HMP has reached the stage where it's starting to generate testable hypotheses.  Momentum isn't flagging.

Compare to the human genome project:  He thinks there won't be a reference microbiome for each site.  Need to learn to distinguish 'healthy' microbiome variation and dysfunctional variation.  Need a multidimensional view (genes, transcripts, proteins, metabolites...)  Very valuable that we have close interactions between clinicians and researchers from the very beginning.

Coming hardware technology:  PacBio, 454 junior, Ion Torrent, MiSeq, SOLiD. 

PacBio: longer reads very valuable, run times shorter, needs less sample.  Example Enterococcus faecalis: ~50x coverage gives ~100% coverage.  Average read length 587 bases, 12% >1 kb.  Simply aligning the >2kb reads covers most of the genome.  Accuracy?  Peak 85% accuracy (= 15% errors) as advertised, verylow accuracy.  But errors are random, not systematic, so deeper sequencing corrects the consensus.  Use for 16S?  High error rate is a problem, but if can get whole 16S genes rather than little sub-sequences, could be very useful.

Computing technology:  continuing methods development.

IHMC1: second Friday morning session

Brandi Cantarel: Crohn's disease insights.  Swedish identical twins - expect similar microbiota from genetic identity and common environment.  CSI metaphor.  454-16S and whole-genome sequancing and meta-proteomics.

Previous data (16S) showed different clustering of bacteria with Crohn's at different colon sites, away form non-Crohn's.  1 healthy pair, 1 both-Crohn's pair, and two discordant pairs (one twin healthy, other with Crohn's).  Now looking at whole-genome sequence data for same twins.

Look at predicted proteins and at proteome...  Lists of categories that were and were not informative...  More genes expressed in healthy than in disease communities.  Two proteins differ significantly:  thiosulfate somethingtransferase and something else (fits previous info about sulfur metabolism in Crohn's).  Crohn's has Bacteroides proteins, healthy have Faecalis???  From genome sequences, not much correlation with enzymes for use of different energy sources, but see some correlation (genes present but not expressed) in proteome.

No correlation with twin phenotype = no effect of genes.  This must be more surprising than she sounded. 

~30% of proteins were from host.  But we don't learn anything (at least not yet).

Sarkis Mazmanian: Unlocking the evolutionary mysteries of microbiome-immune symbiosis.  Experimentomics!

Hypothesis: Probmen in irritable bowel diesases (including Crohn's) is lack of immune tolerance to 'normal' flora, causes 'dysbiosis' = imbalance. 

Lab studies:  Bacteroides fragilis (Gram-neg, obligate anaerobe, makes > 8 capsule types, two have zwitterionic - pos-charge unusual.  Treat model organism with 'polysaccharide A' (capsular polysaccharide of one type), reduce inflammatory response, reduce symptoms.  See PSA (but not PSG  = another capsular type) is packaged into vesicles.  These vesicles protect against colon disease.

Role of type 6 secretion system:  Bacteria can secrete immunomodulators.  If this is disrupted, get overactivation of host T-cell-induced inflammation.

Elhanen Borenstein: Gut microbiome metabolic dependencies and determinants.
 
Two studies:  1. Systems biology.  2.  Metabolic dependencies. Using network theory: oversimplified but very messy.  But easy to create on a large scale, and lots of tools for analyzing topologies.

A.  Reverse ecology:  Predict ecology from genomics.  Identify the set of exogenously acquired compounds (network seeds).  B..Topological markers of environmental adaptation.

Nice ecology-based analysis of evidence for competitive and cooperative interactions in the oral and gut microbiomes.  See coocurrence correlating with competition, not cooperation.  Suggests role of niche-selection, more than species interaction.

Microbiome-wide metabolic modeling produces a network 'hairball'!

Markus Hilty:  The nasopharyngeal microbiome in infants with acute otitis media (AOM) in the era of PCV7 vaccine.  From cultuyre-based studies:  S. pneumoniae, H. influenzae, Moraxella catarrhalis.  All also commensal.


Microbiome study:  healthy vs AOM, antibiotic use, pneumococcal vaccine.  Infants < 2 yr old, nasopharyngal swabs, four winter seasons at a pediatrics hospital in Bern.  Half ind aycare, most vaccinated.  153 with AOM, 10 healthy.  Used 16S PCR, plus culture of stereptococci to distinguish species v. similar by 16S sequences.

AOM microbiome has very low species richness/diversity compared to healthy microbiome.  Total population, mostly Moraxella and Streptococcus, H. influenzae 16%.  Pasteurellaceae absent from healthy infants.  Antibiotic exposure effects: increases Pasteurellaceae!  Effect of PCV7 vaccination:  not much.  Generally lower species richness when S. pneumoniae is present.  No effects of sex, age, or daycare attendance.

IHMC11 Friday first session

Paul Spicer: Indigenous communities and genomics:  lessons for the microbiome.  He's an anthropologist (?) reporting on what the feedback/pushback is from indigenous people.  Bottom line - they're not impressed.

Tribes think scientists are too obsessed with data, not with the needs of the people.  Researchers benefit but the subjects don't.  He recognizes that this isn't really true, but it's the impression we give.  Our apparent difference to human suffering ("we need to bank the DNA before the people vanish...")  ELSI is 'research on research' - of no value to the indigenous communities.  But we're not funded to benefit them, but to get the research done.  We certainly shouldn't claim to offer unreasonable benefits (destroys trust).

We need to develop the capacity for partnerships - build ability of communities to benefit.  Need to work with village schools, not just target PhD and post-doc programs to the communities.  They try to form partnerships with us, to get what they value, but these attempts fail.  We need to find ways to benefit them.


Makedonka Mitreva: The microbiome of healthy humans commonalities and variations.  HMP goals:  sequences of target genomes, unbiased 16S surveys, metagenomics.  She'll report on metagenome analysis.  100 individuals, 2 visits for most, 6 body sites.

Presence of matches to reference genomes (she says 'strains'):  Alignment with 80% identity to reference genomes.  Depending on site, 33-77% of reads find a match.  Complex clustering (based on 'DCPM') by site .  (I don't know what this is supposed to illuminate.  She must be assuming more expertise than I have, as she hasn't said what questions this is supposed to answer.)

Using alignments to reference genomes, the bacteria up your nose don't overlap much with those in your vagina.  Other sites show more overlap.  But how does this change if they instead measure diversity by 16S sequences?  Not too much - there's a good overlap for most sites.  Not so good for dorsal surface of the tongue.

What about genetic variation with an identified taxon ('strain')?  e.g. Streptococcus oralis.  See lots of SVP variation, different alleles in different oral sub-sites.  She doesn't interpret but this must mean that different strains of S. oralis are present in different parts of the mouth.

Metabolic profiling?  Just analysis of the diversity and correlations of the profiles.  Correlations between visits are not very strong.  Now she mentions biology!  What can we learn?  'Differential KOs'?  (Google: Kegg Orthology groups).  Two-component system pathways in Staph.

Larry Forney:  (terrific talk!) The temporal dynamics of the vaginal microbiota in reproductive age women.


Common 'wisdom' about the vagina: Dominated by Lactobacillus, nedcessary for health, restricts growth of baddies by low pH (~4.5), high lactate, other factors.  Transitions through lifespan:  colonized at birth but not studied in childhood.  Puberty:  estrogen production, colonization aand succession.  Menopause:  estrogen down, less glycogen, less lactic acid.

Bacterial vaginosis = 'disturbed microbiota' (?).  Clinical criteria or Gram stain for lactobacillus.  Some women at much higher risk than other.  IS his due to differences in vaginal comminities?

Questions:  what are the communities?  Are they dynamic? What factors determine the community structure?  Are these communities and factors associated with disease?

Study:  410 healthy women, 4 races, 16S 454 sequencing.  In 95% of women, find one of 5 different community types, 4 dominated by different single Lactobacillus species, and one with heterogeneous lactic acid bacteria.  (5% of women don't fit any of these.) But in individuals, the communities were all different, approaching more-or-less to one or two of the five types.

Longitudinal studies over 16 weeks:  Hypotheses?  1.  Each woman's state is stable over time.  2. Each woman's state varies over time.  3 Each woman's state varies but resiliently returns to its core state.  4.  The number of states is limited...(?).

He shows one woman's time course, with blips in community composition corresponding with menstruation.  But only one woman had this.  Another woman: two and a half weeks of very different communities around two menstruation cycles, but stability over other cycles.  And the two changes were not similar to each other!  Other women, dramatic changes irrespective of menstruation or other activities.  He has a movie showing the changes but it vanished from the screen!  

Over the 16 weeks, it looks like most women's communities had a stable core type, with intermittent temporary switches to other types. Vaginal intercourse didn't disturb the communities.  We don't have data about changes over reproductive lifespan.

(Now some heavy ecological diversity statistics...)

Diversity within species:  See changes in ribotypes of the dominant species, even where community type is stable.  So there's quite a lot of variation in the specific strains that are dominant, sometimes correlated with changes in the abundances of other species.


Lactobacillus iners:  very dependent on nutrients from the host and the microbial community (a nutritional network).  The L. iners-type communities are more likely to undergo changes than are other types of communities.

Now doing a daily-sample study of 135 women over 10 weeks, including episodes of vaginosis.

Questions:  childhood (ongoing), pregnancy (little info but becomes more stable), menopause (a few cross-sectional studies, Lactobacillus present still)?  Time scale faster than at other sites?  Diet?  Innate immunity?  Sexual arousal? 

L. Caetano Antunes:  (NOT a terrific talk)  A metabolomic analysis of the mammalian gut microbiota.  Starts with the obvious (host vs pathogens, basic info about microbial communities and omics), he should have cut all of this from his talk for this audience.  Then he wouldn't have to talk so fast!  Bad slides too - tiny text, or blurry images with illegible text.  I bet this is a job talk, speeded up to fit in the short time allowed.

Mice:  collect feces before and after treatment with streptomycin, extract with acetonitrile (accomplishes what?), use mass spec to identify metabolites.  His story is about leukotriene B$ levels, which vary a lot, in response to streptomycin treatment and between individuals.

Finally, the recognition I've been waiting for!

"Rosie Redfield must be the tyrant queen of science."

P. Z. Myers, Pharyngula

Coming up:  More live-blogging from the International Human Microbiome Conference beginning at 8:00 am Pacific time.

Thursday afternoon sessions at IHMC11

Rob Holt: colorectal carcinoma metagenomics.  (His tone of voice makes this sound dull...)  Nearly 20% of global cancer burden known to be due to infectious agents.  They're doing a systematic search for additional agents and factors.

Pipeline:  Start with tumor samples, get RNA, subtract host rRNA, RNA-seq.  (I'm not quite sure what his strategy is...).  Human genome and transcriptome subtracted.  Now look for known viral and bacterial sequences.  What aligns to known bacteria etc, and what potentially novel-organism sequences don't align?

Reconstruction:  could detect 0.1 pg of viral RNA (8 kb) per 2 µg human RNA.  Recovery about 33& of input.  Dull dull...

Why look in colorectal carcinoma? Lots of exposure to microbes, precedent of H. pylori, the guy in the next lab had some we could use.  Has 669 candidate seq (no, candidate bacterial species )from control, 492 from tumor, 250 common to both?  Most reads from the known gut bacteria, some phages.  (Different from Meta HIT (INRA intestinal tract).

High abundance of Fusobacterium nucleatum (anaerobe, Gram-neg, periodontal, appendicitis) in tumors but not controls (86-fold overrep), but no other significant correlations.  Mostly two of the eleven patients, and mostly rRNA sequences of the bacteria.  Look at other tumors, see commonly overrep of Fusobacterium.

"This is quite exciting" (said in monotone).  No clinical correlations found with presence/absence of Fusobacterium, even though this species known to be pathogenic.

Question about tumor site (different types of tumors in different places in the colon and rectum) and about the patients not having their gut 'cleansed' before the tumors are excised (so the tumors could have a lot of surface bacterial contamination).


Heather Maughan:  Ecology and evolution of the cystic fibrosis lung microbiome.  Healthy lungs are not completely sterile, but we can get rid of the bacteria we breathe in.  But in CF lungs the bacteria are like unwelcome house guests, very hard to get rid of.  Follow bacteria in CF lungs through childhood and adults.  See changes in prevalence of the common bacteria.  (H. influenzae important in young patients but not adults.)

Get 'explants' (= lungs removed from CF patients receiving lung transplants) plus hundreds of sputum samples from baseline and exacerbation patients.  Sequencing 16S rDNA.  Also whole-genome microbiome sequencing (includes host DNA, viruses and phage).  Now lots of info about methods of analyzing the data - how many reads, how efficient, appropriate cutoffs, false-positive rates, worst is still good 0.08%.

Individual patiets differ in their microbial diversities.  All have Burkholderia, most Haemophilus, Pseudomonas.  Peptostreptococcus (anaerobic),  Some reads mapped to chloroplast relatives - patients may be breathing in pollen.. 



Kristine Wylie: Human virome in children and adults.  (Oh-oh, nasty audio problem.  There's an echo from the rear speakers, making it very hard to understand what she's saying.  At the question period, discover the same problem with audience microphones audio.)

Children with fevers have more viruses in the nasopharynx, only febrile children have viruses circulating in plasma (except anello?).  Anello virus in plasma only in afebrile children.  These viruses have only recently been discovered,  they're common and genetically diverse, and have not yet been associated with disease.  Bocavirus.  RSV.  Human rhinovirus QPM.  Human parainfluenza virus.

Human microbiome project samples:  Two times from same person, from multiple body sites.  Find the same kinds of viruses at many sites.  Find papillomaviruses in various sites.



Frederick Bushman:  The virome of the human gut: metagenomic analysis of changes associated with diet.

Cross-sectional study of diet and stool microbiome (COMBO)  See strong associations between certain bacteria and certain dietary components.  Patterns of phyla of bacteria with high-fat or high-fiber, and of specific bacteria.  But dietary effects accounted for only a small fraction of total variation between individuals.

Controlled feeding experiment (CAFE)  10 healthy volunteers, captive feeding in a hospital for 10 days; low fat or high calorie diets.  Variation much lower between individuals, except on day 1.

VIROME:  Goal is to characterize the whole viral population.  (Lambda control gave quantitative recovery.)  10^10 phage per gram of stool!  Circular contigs (genomes?) all about 5-6 kb.  Linear ones very diverse lengths.  7000 new virus genomes!  19-785 per individual sample.  Lots of 'unknown'!  No contigs of eukaryotic viruses at all, but bits of eukaryotic viral genomes in phage genomes.  He thinks there has been lots of misidentification - what appears to be DNA indicating presence of a eukaryotic virus is really jsut a bit of phage genome.  See CRISPR system used to compete with other phage (I forget what CRISPR does).

Lysogeny:  He thinks they found lots of evidence, but the audio problems mean that I'm not sure what he said.

Interpersonal variation in virome:  Most variation is accounted for by who is being examined.  Subjects on different days are more similar to themselves than they are to other subjects on the same day.  But subjects on the same diet do get more similar than they were before.

What determines phage abundance?  Abundance of host?  Lysogenic induction?  No evidence for consistent relationships to putative host abundance.

This is a big meeting

IHMC11 last Wednesday session:

Elodie Ghedin: the lung microbiota in SHIV-infected macaques.  (Alison Morris is clinical side of this work, Elodie is bioinformatics)

An animal model of HIV-associated lung diseases; work in ppreparation for studies of human HIV lung microbiota.  Lung disease is more common in HIV people with COPD (chronic obstructive pulmonary disease) on antiretroviral therapy than on people not getting drug therapy. Independent of smoking, HIV reduces pulmonary function, increases Pneumocyctis infection which increases obstruction.  Collected bronchial lavage samples monthly (15 months), also blood, CD4 counts, clinical measurements of function.

Results very different for different monkeys!  (even though one cohort in a closed facility)  Follow CD4 T-cell counts do decline at weeks 4-5, fairly consistently, and then rebound.  If look at airway obstruction, see decline in lung function at weeks 9-25 in about half the monkeys.  She says the data is 'entropic' - a slightly evasive way of saying that it's noisy.

Pneumocystis:  in 10 of the 12 monkeys in this experiment, but have another set that didn't get Pneumocystis.  No significant difference in microbial communities of monkeys with and without Pneumosystis, but big difference to monkeys not infected with SHIV.  (Same in a different set of monkeys.)

I had thought that the lungs of healthy people were close to sterile, but my expert seatmate twells me this is now known to be wrong.  She says that lots of bacteria live in healthy lungs.


Charlie Xiang: Diversity of bacterial vaginosis.  (Oh dear, strong accent - luckily he's reading the text on his slides, but needing to look at the slides makes it hard to write.  And so far he's just giving a boilerplate talk, telling this specialized audience what he would tell a non-expert audience.)  He heads China's 973 program:

Part 1:  This project: 50 women with bacterial vaginosis and 50 healthy controls.  16S rRNA-PCR-DGGE (not very informative, why bother?), 454 sequencing, qPCR.  Two ecological diversity indices, one says vaginosis microbiota is more diverse, other says it's less diverse.  But looking at the taxa, vaginosis communities look much more diverse.  (He's showing the figures from a paper (Ling et al. 2010 BMC Genomics 11:488-503), not optimized for presentation in a talk.)

Part 2:  Women with bacterial vaginosis, 22 with cervicitis, 18 with both, plus healthy controls.  Total numbers of bacteria similar, but identities different.  But no cervicitis-specific pattern.

Infectogenomics!

Questions:  Might vaginosis be simply due to a deficiency of Lactobacillus, rather than to increased numbers of other species?  Or to some decrease in innate defenses of the host?


Alan Walker: Human colon vs diet.  'Non-digestible carbohydrates' (non-digestible by humans) are major sources of nutrients for colon bacteria (!10^11/gm of colon contents).  e.g. plant cell-wall polysaccharides.  Different bacteria do different steps.  Butyrate is a major energy source for many gut species.

High bacterial growth rate in proximal colon (on CHO), in distal colon bacteria switch to protein fermentation (byproducts more toxic for human cells, may be major cause of colon cancer).  We know little about effects of ~normal diet on colon bacteria (not probiotics or weird supplements).  People are not very honest about reporting what they eat.  So they locked overweight/obese people up and controlled everything they could eat for 10 weeks!  Collected fecal samples.  Hard to get people to eat a lot of bran - 'resistant starch' is easier.

Microbiota?  DGGE, 16S sequencing, qPCR: Samples clustered by individual, not by diet.  Saw only a few changes that correlated with diet (only significant for resistant starch).  Ruminococcus up in most people with resistant starch; in other people different bacteria went up during diets high in resistant starch.  Now he's analyzing whether the bacteria are in the liquid fraction or the feces or attached to the fibers, but I missed hearing why this is interesting. 

Bottom line:  different people's bacteria respond differently, and responses by different bacteria have different consequences.  Gee, just like ecology, every ecosystem is different!

Question:  Enterotypes?  He's not convinced this classification is useful.


George Weinstock reporting on work with Martin Antonio: Microbial ecology of the infant nasopharynx: impact of the PCV-7 vaccination in Gambia (West Africa).  MRC has labs and field study sites in Gambia - very good for field studies.  (George is stepping in because Martin had visa problems - George hasn't seen these slides!)

Child pneumonia - 70% of deaths in Africa and Asia.  40% are pneumocccal (S. pneumoniae) - this is the meningitis belt.  Lots of serotypes, some virulent (rarely carriage).  New vaccines include the major African serotypes.  Are the vaccines changing the abundances of different serotypes?  Will serotype switching happen?

Aims: what is normal nasopharynx microbiome in first year of life, and does the vaccine change this?

3 groups, ~100 total.  1=Control = non-vaccinated infants from non-vaccinated villages.  2= vaccinated infants in non-vaccinated villages; 3= vaccinated infants in vaccinated villages.  17 samples from each infant in first year of life.  Sequence 16S rRNA with 454.   

Haemophilus (my organism) is abundant in all of them pre-vaccination (10-40%), but declines after 27 weeks and isn't changed significantly by the pneumoccal vaccination.  In response to a question, he said that all the infants received the Haemophilus vaccine (Hib?).

See significant effect of vaccination on some species.  Diversity measures - nothing exciting.

IHMC11 Wednesday afternoon session on ethics

The session's topic is 'ELSI', which I'm told is an acronym for ethical and philosophical issues...  The moderator repeated the acronym but didn't explain it.  (Later, ELSI=Ethical, Legal and Social Issues.)

Amy McGuire:  Identifiability of the human microbiome - Investigator perspectives.  Data sharing and identifiability (sounds dull but turns out to be really interesting, but she talks too fast!)

Policies favour data sharing and accessibility, developed out of large-scale consortium projects developing a community resource.

Problem:  It used to be enough to strip the personal identifiers from the sequence file and associated phenotype information, but now we realize that DNA sequences are themselves personal identifiers (30-80 SNP diplotypes are enough to identify anyone).

Now. dbGap- data is behind a firewall, includes phenotypic information but is only accessible by 'professionals'.  (Professionals are trusted with confidential information.)  Even aggregated sequence data is now being moved behind professional barriers.  Is this overkill?

HMP data should be public but personally identifiable data should be put at dbGap.  Effort to remove human sequences from the data sets.   Also collect human DNA form a blood sample - aggregate data is released but individual sequence data into dbGap.

Even the cleaned up microbiome sequence data contains some bits of human sequence missed by the clean-up screen against human genome.  Alternative would be to screen against bacterial genomes, but risk losing unknown bacteria.  Researchers asked, is this overkill, given that participants have signed consent forms and the risks of harm to a participant are very small?  Is it a waste of $ and resources?

Another, more philosophical issue, is whether your microbiome is part of you - do participants think their microbiome is apart from them?  Will they continue to think this?  And will our microbiomes come to be ways to identify us personally, as our DNA sequences now are?  (A test using bacteria on our keyboards - we leave our bacterial 'fingerprints' on our keyboards!)

Lots of concern about the scientific value of linking microbiome data to metadata about the person it's from, and about how concerns for privacy could compromise this.  Participants have a lot of trust of the researchers, and are quite comfortable with data sharing about clinical information (not names and addresses).  They care more about feeling respected as a contributor that about protecting their privacy.

Need strong penalties for misuse of the data.  Rather than compromising the science, put in place strong deterrents to misuse.

Keiran O'Doherty:  Social and philosophical ramifications of HMP.  Ahhh, he speaks much slower...  But his slide background has streaks that simulate failing to block out all the sunlight in the room!  He's reading a quote from a philosopher.

The main point:  Do we think our bacteria are part of us?  Does changing my microbiome change me?  My questions:  Does brushing my teeth change me?  What about getting a dental crown, or dentures?  My hair and fingernails are part of me, but does cutting them change me?  Yes but No.  We're comfortable with having the external bits of our bodies changed, and don't think that this change who we are.   He's talking about genetic engineering of embryos and about eugenics, to set a baseline.

Scientific questions:  How heritable and stable is a microbiome?  Is there a 'core' microbiome? Is there horizontal transfer of microbiomes between people?  When we modify an individual's microbiome, are we tampering with the human microbiome-poo, (like eugenics and the gene pool?  (Yikes, now he's talking much faster...)  Do we need to worry about preserving human microbiome biodiversity?

Could we do good by modifying microbiomes?  Good at the public level.  the microbial equivalent of fluoridating the water supply?

Bottom line:  we should be cautious.


Laura Achenbaum:  HMP ethics from the trenches.  She's part of Amy McGuire's group (speaker 1).  Now she's telling us stuff we already know, and details about their survey population (OK, that is what the talk is about).

The consent form is 15 pages - do people really read it?  The participants say it's OK, it reassures them that all the issues have been taken into consideration?  Can the participants really understand it?  Hard to say because so far they've only tested it on a med-school derived sample.

Trust is a big factor - the government/NIH is protecting me...(nasty stretched image of clasped hands).  (nasty stretched image of apples).  Participants liked the financial compensation, and appreciated that they were contributing to socially valued research.


Mark Sagoff: Ecological metaphors.  More philosophical - from interactionist credo (self interacting with environment, e.g. norm or reaction, genotype vs environment) to ecological credo (whatever that is).  He's clearly steeped in the philosophy of ecology.

(Oh dear, he's a bit rambling and he's just put up a slide full of text, and he's quoting the same source as speaker 2.)  He's pointing out how ecological terms are used  he seems to think these are metaphors, but I think they're the reality.  Microbiome research is inheriting a history of confused nature- nurture controversy and a muddle of ecological ideas.

In the past, ecologists have used the human body as a metaphor, but now human biologists are using ecology as a metaphor.

The 'individualist' concept in ecology.  ('individual' meaning, in this context, 'species'!).  Some ecologists now think there are no patterns ad principles in ecology of communities.  Instead any particular community is just whatever assemblage of species that live there.

Bottom line:  The science of ecology cannot offer a conceptual hand to microbiome research, nor a normative analogy.  Because ecology traditionally excludes humans.  He looks for comfort in Aristotle:  material cause - the molecules we're made of; (whoa, now he's gone weird...)  efficient cause: how microbiome causes differences in humans; formal cause:  oops, lost what this is and the last kind of cause.

I asked him, just to confirm:  he says yes, it's true that microbiome research is indeed ecology, but ecology won't help us understand the microbiome because ecology itself is theoretically bankrupt.  He says that humans should now be seen as the model system for ecology.

Live-blogging the Human Microbiome meeting, first morning

Nina Lin: (also poster #67) Looking for symbiotic interactions using micro-droplets of medium (~1500 per cm2?). Using E. coli auxotrophs in various combinations. Media and solutes don’t diffuse between droplets (though gas could).

Now try murine fecal microbiota (just getting started). I missed the beginning – don’t know if anaerobic, nor culture medium. Now trying to expand culture conditions, controlling oxygen in defined gradient.

TRFLP: amplify 16S with fluorophore tag, cut with restriction enzyme, run in capillary electrophoresis. See much less diversity with pooled 800 drops and less with pooled 100 drops. But need control of entire population cultivated under the same conditions.

I don’t know whether they can retrieve bacteria from individual droplets.

Michiel Kleerebezen:  Microbiota of the small intestine are little-known because so hard to sample.  Use people with ileocolostomy to access the small intestine.  Can sample repeatedly, examine effects of dietary changes.  Also volunteers who swallow a catheter for sampling.

Uses a human intestinal tract chip (HIT chip) to identify bacteria.  Ileostomy effluent comes from the terminal portion of the small intestine, but its bacteria resemble that of the proximal ileum (quite a ways upstream).  Now doing a metagenome and metetranscriptome analysis.  Compare to large intesting data.

Metatranscriptome:  Simple CHO utilization, fermentation.  PTS transport very active.  Biotin synthesis very active.  Who's doing what?  PTS exclusively streptococci and E coli and relatives.  Hypothesize that lactate and acetate are excreted and used by other bacteria.  Biotin made by everyone.

Who's there?  Not many lactobacilli (yogurt) but if feed them a probiotic (10^9) lactobacilli see a spike to 50% lactobacilli in effluent over a day.  How much did the lactobacilli grow?  He didn't say - maybe I can ask later.

Study in 8 volunteers, crossover design.  Fast, then drink every 30 min for 6 hr.  Sample mucosa at 6 hr to see how host gut cells have changed their gene expression in response to the presence of the bacteria.  (Lactobacillus acidophilus, L. casei, L. rhamnosus, each at 2-5 x 10^10.) 

Showed 500-1000 differentially expressed host genes.  Different gene-category effects by different bacteria.  BUT sample size is small (8) and there are no error bars.  He wisely claims to only have a "hypothesis-generating model".

Curtis Huttenhower (from Human Microbiome Project (HMP)  Responsible for 30 posters at this meeting!): Overview:  300 people, 15-18 body sites, all being sequenced in different ways at various centres.  Big questions:  Who's there?  What are they doing?

What to do with your (you the researcher) metagenome (from whole-genome shotgun sequencing)?  Meed massive computational methods, to convert pinpricks of data into a big picture.  He's telling us about metabolic reconstruction.  Use cleanup analysis to convert sequence reads into measures of gene abundance, and then to smooth these into estimates of what pathways are present (encoded in present genes).  Remove pathways only done by organisms absent from the sample.  (Uses analysis methods taken from linguistics analysis!)  The whole analysis was validated by sequencing and analyzing synthetic communities.

Output:  Gene presence/absence.  Pathway present/absence.  Then pathway abundance in the sample. 

Data from 741 HMP samples:  Lots of variation in which bacteria are present in different individuals, but much less variation in which pathways are present.  Zero SPECIES of bacteria area present in all samples.  But 19 PATHWAYS are present in all samples.  Pathway abundances are consistent across individuals but variable across environments (body sites).

Methods from ecology.  'Alpha diversity'

To what extent do the differences in variability between 'Who's there?' (very variable) and "What are they doing?" (not variable) result from the nature of what is being measured rather than from ecological properties?

Look at pathways 'essential for life' and baxic metabolism.  These are (unsurprisingly) very stable.  If you zoom in on the metabolic details, see more variation in how specific pathways are being implemented.

Vaginal microbiome vignette:  Very few genera present, mainly a few reference strains (Lactobacilli, Gardnerella).  This affects the metabolic reconstruction.

What a real astrobiologist at NASA has to say

I'm posting here a comment from Lynn Rothschild, a genuine astrobiologist at NASA (she's also at Brown):

Rosie,

Great job once again. However, between this and the recent arsenic paper fiasco I feel obliged to speak in defense of NASA. NASA is not a monolith but more like a large university with ten or so campuses. Richard Hoover is an engineer at Marshall Space Flight Center, and not a biologist by training. In fact, there are professional microbiologists at Marshall conducting ISS monitoring but I don't believe they were in any way involved with this work. From what I can tell, all of Mr. Hoover's assertions about life in meteorites are in non-peer reviewed journals and that his awards are in engineering. His papers may not have been approved by the Agency prior to release, and even if so, there is no guarantee that they were reviewed by biologists, micropaleontologists, or other relevant professionals.

In contrast, the arsenic paper was not authored by NASA scientists at all but rather was supported by NASA funding, a distinction not made in most press reports. The specific conclusion of the paper is not an official NASA position that we have all bought into. We are not told to give a party line on any scientific discovery just as presumably you are not by your institution.

Bottom line: there ARE many distinguished scientists at NASA, including some microbiologists, one or two evolutionary biologists, geologists, and astrobiologists. But I for one would prefer to go the traditional peer-review route rather than science by weekend blogs. I may not go viral that way, but in the long run, it is the right way to conduct science. So please, don't tar all of us with the same brush as you would hope others would not besmirch your reputation because of the papers of others in your institution.

But that begs the question of why all the excitement and passion with this particular story? Because the the search for life elsewhere is one of THE most compelling questions in biology, and surely the discovery of extra-terrestrial life will revolutionize our understanding of the origin and evolution of life. Of course a few well-placed stories (again note by non-biologists or geologists) and twitters on Friday night is really what made this story go viral.

Is this claim of bacteria in a meteorite any better than the 1996 one?

A new paper from a NASA scientist claims to present evidence for bacteria-like organisms in some meteorites.  (Richard Hoover, 2011, Fossils of cyanobacteria in C11 carbonaceous meteorites. Journal of Cosmology 2011, vol 13.)

I don't know much about meteorites, but here's my evaluation: (Executive Summary: Move along folks, there's nothing to see here.)

What the author did:

He fractured tiny comet-derived meteorites (0.1 - 0.6 g) from two events and examined the freshly broken surfaces.  He claims to have observed structures that are remnants of cyanobacteria.


These meteorites are of a special very rare type (only 9 are known).  They are about 20% water, and soft enough to cut with a knife.  They mainly consist of minerals cemented together with magnesium sulfate ('Epsom salts'). They come from asteroids and comets, not planets like the Alan Hills meteorite from Mars.  Hooper's reasoning that they come mainly from comets seems reasonable to me.

They contain quite a bit of organic (carbon-based) material, but I don't know if this differs significantly from the polycyclic aromatic hydrocarbons known to be present in comets.  It's true that PAHs found on Earth are usually biological in origin (think of the tarry crud that accumulates on your barbeque grill), but that doesn't mean that PAHs from space have biological origins.

An important concern with this kind of study is contamination with terrestrial organisms before examination.  He doesn't say how the meteorites have been stored before he obtained them, nor how the surfaces of the meteorites were treated before being fractured and examined.  He doesn't say how they were fractured - might they have been cut with a scalpel blade or just pressed on until they crumbled?  He says that the tools were flame-sterilized, but not what the tools were or how they were used. 

He used two examination techniques.  FESEM is field emission scanning electron microscopy - this seems to be a higher-resolution form of scanning electron microscopy (SEM), with the usual risks of artefacts.  The fractured surfaces were not coated with anything before being analyzed - I don't know what effect this might have.  The other technique is energy-dispersive X-ray analysis - I gather that this is an add-on to SEM that can scan a specimen and report on the abundance of specific atoms at different positions.  Its results can be reported as the distribution of atoms at a particular position or as an image of the specimen, shaded to show the varying density of a particular atom.

Results. 

He shows an image and analysis of one filament from the Ivuna meteorite.  It has more carbon than the surrounding material but no detectable nitrogen or phosphorus. 

He bolsters his claim that it's a bacterium by showing an image of the giant bacterium Titanospirillum and an image of another filament from the meteorite.  His claim that the sulfur granules in this second   filament are like those of Titanospirilum is weakened by the very high sulfur in the surrounding material.  And although this filament is similar in size and shape to Titanospirillum (upper images), the other filament is about 15 times smaller (bottom images, adjusted to approximately the same scale).

The image he shows of an inner surface of the Orgueil meteorite has more filaments (no attempt is made at quantitation).  These are more complex in structure and fairly similar to each other, suggesting that they were formed by a single kind of process.

The atomic analysis is not at all convincing.  He claims that different parts of the filament have different composition, but doesn't present any control analysis of the variability of the measurements or of the background values for positions away from the filaments.  He claims that the atom-density scans show enrichment of carbon and oxygen in the filaments, but this looks very weak to me - the only strong signals are for magnesium and sulfur.  Again there is no detectable nitrogen or phosphorus.

He spends a lot of text discussing the morpohlogical similarities of these filaments to cyanobacteria, but I don't regard these similarities as worth anything.  Filamentous bacteria are very morphologically diverse, and additional variations in appearance are likely to result from inconsistent preparation for electron microscopy.  It's probably pretty easy to find a bacterial image that resembles any fibrous structure.  In the absence of any statistical evidence to the contrary,  it's prudent to assume that such similarities are purely coincidental.

The author tacks on quite a bit of other less-than-compelling information intended to support his claim that life from space is plausible.  For example, he shows photos of colonies of coloured microorganisms to support his argument that the colours seen on the surfaces of Europa and Enceladus are biological in origin.

Bottom line:

The Ivuna meteorite sample showed a couple of micron-scale squiggles, one of which contained about 2.5-fold more carbon than the background.  One of the five Orguil samples had at least one patch of clustered fibers; these contained more sulfur and magnesium than the background, and less silicon.  As evidence for life this is pathetic, no better than that presented by McKay's group for the ALH84001 Martian meteorite in 1996.

The Journal and the Editor aren't very impressive either:

The journal proudly announces that it is obtaining and will publish 100 post-publication reviews.  But did it bother getting any pre-publication reviews?  It will be shutting down in a few months, after only two years of on-line publication (the 13 'volumes' are really just 13 issues).  Its presentation standards are pretty bad - there doesn't seem to have been any effort at copy-editing or formatting the text for publication (not even any page numbers). 

Chandra Wickramasinghe is the journal's Executive Editor for Astrobiology, and presumably is the Editor responsible for this article.  I heard him give a talk pushing panspermia about 10 years ago (the audience was an undergraduate science society at Oxford).  The talk was very slick but dreadfully bad as science.  The evidence he cited to support his arguments wasn't actually untrue, but he twisted everything to make his arguments seem stronger than they were.  He argued like a lawyer - his only goal seemed to be convincing the audience that his conclusion was correct, regardless of the contrary evidence that an unbiased consideration of the evidence would provide.  Thus I wouldn't trust his scientific judgment about anything concerning astrobiology.


Hoover, R. B. (2011). fossils of cyanobacteria in C11 carbonaceous meteorites Journal of Cosmology 13.

Microbiome meeting poster

The postdoc and I are both doing posters for the International Human Microbiome Conference next week.  His is about his work on uptake specificity and mine is about his work on recombination tracts.

Because we've finished the manuscript on recombination tracts and submitted it, I'll be able to use the figures he prepared for my poster.  But I'm working on a new figure for the Background, one showing a mucosal epithelium with lots of bacteria and DNA in the mucus layer.  There will be several different shapes and colours of bacteria, in microcolonies and dispersed, and DNA fragments from all of them against a background of abundant human DNA.

I'll post it here when it's done (later today - the posters have to be finished by Monday so there's time to get them printed).  The postdoc could use it in his poster too if he likes.