Saturday, 23 May 2015

Guinea EVD cases rise - but not like they have before...

Edited for clarity 24MAY2015 AEST
The uptick in cases this week from, in particular, Guinea interrupted what was looking promisingly like a continuous downward trend in cases all the way to zero - yes, that was too much to hope for.

However, it's worth keeping some context around this:
  1. Everyone actually working in this space has forewarned us that getting to zero Ebola virus disease (EVD) cases was never going to be an easy journey.
  2. The indicators have consistently shown more reluctance in Guinea to "kick out Ebola" than in Liberia or Sierra Leone. What caused that reluctance, I don't fully understand from my totally uninvolved chair a million  miles away.
    I know right? Surprising.
    Yes-I'd like a specific reason, all wrapped up and presented to me. I'm simplistic and selfish that way. Get that for me will you?? 'Cause I'm totes sure you haven't been trying your collective butts off all this time.
  3. There were fewer new confirmed EVD cases this past week than in the tally for the week before or for other weeks - look back at January 2015, or October 2014 or June 2014, or any of a number of other dates when cases were accruing at a much faster rate than now (graph below).
    That is a silver lining. It's far from ideal, but it's not a return to the worst of it.
All along there has been something different about Guinea and that is now a clear sticking point in the final push to rid its people of Ebola virus|Makona. It never reported, as Sierra Leone did, days with averages of more than 100 cases. 

If we knew what was different about the people, communication, geography, weather, traditions, habits, thinking, ETUs, labs, government...or whatever..then we could perhaps better target the problem(s) and get to zero sooner. 

That will be core business for the next step to occur.

Click on graph to enlarge.

A good week for viruses...not so great for humans...

Edited for clarity 25MAY2015
Middle East respiratory syndrome coronavirus (MERS-CoV) managed to get out for some sightseeing - travelling to South Korea this week - and Ebola virus|Makona was given a helping hand to spread to new people in Guinea and Sierra Leone with a small splurge of new confirmed cases.

MERS has now trickled into 24 countries world wide as shown in the European Centre for Disease Prevention and Control's (ECDC) epic 'travel-by-plane' map.

Media preview
The original of this is created by the ECDC and is presented here.
Click on image to enlarge.
Meanwhile, a crude extrapolation from current Ebola virus disease (EVD) case numbers saw the predicted date when we might reach zero cases, move further into June. 

This could pull back again or it could move further away if the EVD clusters and sporadic cases continue to spread. We can't model that because it's entirely down to unpredictable human variables. We can list what those are, we can better prepare for them, we can educate about them and how to prevent them and we can acknowledge that they are real, but we cannot know when and in what mix they will come into play.

Extrapolation of the public data for confirmed Ebola virus disease cases from
WHO. The most recent week is boxed in red and bucked the trend of declining
 cases. To see how I made this please visit here.
Click on image to enlarge. 
The newest EVD cases remain mostly clustered around the Forecariah prefecture of western Guinea, on the north west border with Sierra Leone but also 5 new cases appeared in the north west of Guinea in Boke prefecture, which borders Guinea-Bissau. 

Geographical distribution of new and total confirmed cases
From the World Health Organization's Ebola virus disease Situation Report, 20MAY2015.
Click on image to enlarge.
Since the last EVD SitRep, two days of reporting have seen fewer cases than in the same two days of the week before. 

So there's that. 

Quickly reporting what is actually happening is invaluable for all sorts of reasons. Modelling and prediction allow us to get ahead of the virus. But having the data, and having them available publicly remains a challenge for every country and for every outbreak. 

Public health data are about the public's health. If it has been considered worth collecting and collating, why not communicate it too?

Thursday, 21 May 2015

MERS-CoV jumps a flight to South Korea...but from where?

It could be Qatar, Bahrain, the United Arab Emirates (UAE) or the Kingdom of Saudi Arabia (KSA). Any of these may have been the country of origin for the infected person who returned with a bunch of microscopic passengers, to the 24th country to host a case of Middle East respiratory syndrome coronavirus (MERS-CoV) infection - South Korea

The infected man then passed the spiky parasites on to his 63-year old wife and to a 76-year old man with whom he shared his hospital room. Close contact. From what we know of the MERS-CoV - it's a pretty ineffective transmitter among us humans types, preferring instead to give the hump to dromedaries.

Qatar seem less likely as it appears to have only been an airport transit point. If it's Bahrain, then we have 25 countries as Bahrain has not yet reported a MERS-CoV positive person. Both the people and the camels of the UAE and KSA are well known to this virus both in humans and camels. 

We await the clarity of the World Health Organization's analysis in a Disease Outbreak News (DON) article - although this might be a tough one to unravel.

Click on image to enlarge.

Thursday, 14 May 2015

Liberia gave Ebola the boot...and a virus may soon be removed from the wild

The people of Liberia have earned our respect, some time for national celebrations and frankly any other rewards that may flow from denying the Makona variant of Ebola virus any hosts among their community. 

The world considered this viral species to be one of the list-toppers when it came to ranking the causes of the most scary acute infectious diseases. Ebola virus has been the basis for all sorts of 'end-of 'the-world' mutating virus horror movies, books, and TV shows. It's not at all surprising that the public view of an Ebola virus infection had long been one of blood, fear and terror.

Figure 1. The decline of the Makona variant of
Ebola virus in Guinea, Sierra Leone and Liberia
(now free of EVD transmission).
Click on image to enlarge.
Behavioural change was a major factor in reducing virus transmission in Liberia. Alongside that was a broad range of aid given from within and beyond Africa's nations. By working together, a widespread outbreak that was not initially thought likely to happen at all, was routed. 

For now. 

Liberia is not immune to new cases of Ebola virus disease (EVD) crossing its borders or popping up due to a new animal-to-human jump (a zoonosis). That could happen any day - it might be happening now. But those who are still on watch will be searching out new cases while the remaining sites of transmission - Guinea and Liberia - do their best to deny Ebola virus a chance to replicate and spread. The people of Liberia will keep watch help because they have learned very tough lessons about viruses, epidemiology and communication. At least 10,604 suspect, probable and confirmed EVD cases, 4,769 deaths and way too many stories of sadness and families destroyed are a very strict teacher. 

Figure 2. The number of confirmed EVD
cases (yellow) grinds to a standstill. Only
9 cases in the week to 10th May 2015.
Click on image to enlarge.
The crude prediction in Figure 1 suggests that zero cases across all three countries could happen at the end of May, but many stars must align for that to be a real event. 

Human factors - the causal and sustaining variables of any outbreak of infectious disease in humans and sometimes animals - remain very much in play. But once that tri-country zero case value is attained, we have 42 days of watching and waiting - from the time the final case tests negative. 

New cases may arise from sources as-yet-unknown. But even if they do keep popping up, it seems very unlikely that widespread transmission will amplify to earlier levels (see the steep slopes in Figure 2) unless a major lapse in attention occurs. Hence,the need for continued vigilance - and Liberia remains on alert for a further 90 days. That more recent figure comes about because we know that infectious Ebola virus can persist in some body sites for many weeks after signs of disease have passed. Whether that virus reservoir is present in every person and whether it actually does cause new Ebola virus infections remain unproven. When you consider what can happen when one person gets infected by an Ebola virus in a tiny remote village in a country that is ill prepared to cope with it and has traditions that lend themselves to its spread...even minor risks rightly come under more intense scrutiny.

What next for this particular virus though? The only place where the Makona variant of this member of the Zaire ebolavirus species will soon exist, is in the freezer of (hopefully) very biosecure laboratories in the US, UK, Africa, Russia, China and probably other laboratories in countries that hosted, evacuated or repatriated cases of EVD. 

There is no sign at all - and this is because of the continued efforts and focus of many currently working throughout west Africa - of the fabled "endemic Ebola" becoming a reality. Unless you mean enzootic 'Ebola'- in which case , it already is, I suspect. It seems very, very likely that the forests of west Africa continue to shelter animal hosts with less mutated versions of this and other ebolaviruses (and filoviruses and who-knows-what else). The host species and route(s) of transmission to humans are yet to be confirmed but for now, we are not too far off eradicating one unwanted viral scourge from the wild. Impressive what we can do when we pull together.

Outbreak resources: more expert detail presented simply, to more people, at a trusted site, quickly, and for free...

Many, many of us have learned a lot about Ebola virus and Ebola virus disease (EVD) over the past 61 weeks - some more than others. 

Some have paid very dearly for their new knowledge and some few have leveraged the event to try and make a buck or draw more attention to themselves or their trade.

Many have been scared - few outside Guinea, Sierra Leone and Liberia have had a real need to be - but fear of this tiny killer is understandable. I stand by my comments on that from back in August when the United States woke up to what had been happening in west Africa for five months, and promptly started freaking out...without evidence of any widespread threat or danger.

Not everyone has a library on everything
For all of the unwanted, unnecessary and often inflammatory commentary, hypotheses, guesses and conspiracy theories, there was some good information to be found about EVD. Sometimes it was only able to be found by academics or others with access to journals that sit behind fee-for-view virtual walls (paywalls). Sometimes the science was too dense for the public to follow - even when they could access it. But most of the time it just took far more digging to unearth the basics than it should have. It would have been good if more of those who could access and interpret that information, had proactively done so.

EVD in west Africa helped generate a lot of publicly accessible descriptive information about some of the technical language of infectious disease outbreaks. But there could be more. New information for public consumption should be...

  • Clear, simple information that can be easily read and shared using today's short, punchy and graphic-laden social media communication tools
  • Information that is quick and easily found or can be found using (way more) friendly search engines. A page of 2,000+ poorly descriptive results returned from a keyword not helpful
  • Broad descriptions about broad topics - not just narrow descriptions for one aspect of one outbreak caused by one virus. We need to explain the wider patterns that are shared among many outbreaks and by many viruses. Ebola virus is not the first bloodborne virus, not the first sexually transmitted virus, not the first virus to spread in vomit and faces or by droplets, or to survive on surfaces, or to mutate, or to have an RNA genome, or to be detected by RT-PCR, or to have its genome sequenced, or to be the trigger for contact tracing, or to have just appeared in west Africa in 2014...etc. Start tying these patterns together to give the public a better sense of what we live with every day, instead of responding to the now and the scary.
  • A single online, well formatted (for multiple devices) site that hosts all this information provided, checked, updated and agreed upon by experts in the fields, written by communicators and hosted by the new and improved World Health Organization (WHO). The world needs a one-stop outbreak info shop that it can rely on. And that shop should be staffed by assistants who are available to answer questions or direct customers to the aisles best suited to their needs. We expect access to information and answers to questions from our phone company, so why not from our World's health experts?
  • Using better citation to acknowledge the reference material in public health information - what is so wrong with letting everyone know what the guidelines are guided by? Anecdotal is not enough.
  • Date stamped to make it clear when it was written and when it was updated. Am I looking at contemporary thinking - or something from 2 days ago before that major discovery/event changed everything we knew about virus X? 
Many public health entities already create pages upon pages of information on each outbreak but some of that is written for people who like to read...a lot...and is in a style that is sometimes too dense and dry with words and phrases that are not well defined. A glossary might also be of use. 

There will always be a portion of the public who seek their news and detail from the loudest and most garish 'news' source. There are also many who would like to be the smartest person at water cooler - but not if that comes at the expense of trying to locate and then wade through reams of technical guff. 

More expert detail, simply presented, to more people, from a trusted site, quickly and for free.

The next 'Ebola' might have a much harder time getting traction in a territory if its population is ready for it, or can get up to speed quickly.

Friday, 8 May 2015

iBola and the speed of research reporting during outbreaks and epidemics...

My thanks to Dr Judy Stone who pointed out the New York Times Health article.[1]

Ebola virus takes advantage of privileged sites

The New England Journal of Medicine (NEJM) has been one of the greatest publishers of high quality clinical information throughout the Ebola virus disease (EVD) epidemic in 2014. 

NEJM has just published an article on a 43 year old male physician who was evacuated from Sierra Leone after an EVD diagnosis, for treatment at the Emory University Hospital Serious Communicable Disease Unit (SCDU). But 14 weeks after his EVD diagnosis, he developed eyesight and hearing problems and was found to have infectious Ebola virus, and viral RNA...In. His. Eyeball.[2] No viral RNA was found on his conjunctiva or his tears. He hearing loss was on the same side as his declining vision.[1] 
An approximate timeline of the Emory patient.
Coloured balloons with vertical lines indicate specific dates.
The absence of a vertical line indicates a general time frame only.
Click on figure to enlarge.
Emory has treated at least four EVD cases evacuated to the United States [17] and has done so with exceptional care and enviable professionalism.[19] They produced a protocols website that became a beacon to the world's public health efforts to prepare for an imported case of EVD.[18] 

This latest finding may go a long way towards explaining some of the issues suffered by survivors of EVD - headaches, eyesight and hearing issues, muscle and joint pain, tingling or other sensations and extreme weakness. Some of these symptoms have affected 50% of survivors by some counts.[4] 

Eyesight problems may not be a long term effect from damage during the earlier infection, but the result of a persistent infection - active virus replication at body sites like the eyeball, that have immunological (e.g. expression go anti-inflammatory chemicals) and physical methods (e.g. cellular barriers) of keeping out molecules, viruses and bacteria, but not necessarily host cells.[5,6,7] Such sites do not usually have a strong inflammatory immune response to pathogens - perhaps to avoid the collateral damage that occurs during inflammation. Should a pathogen get into one of these sites, perhaps by riding in via an immune cell it infected, it can set up house for an indeterminate amount of time and with an unquantified spectrum of damage - from direct viral cell deaths to inflammation. It may also be difficult to get a drug into these sites because of the very discriminatory barriers mentioned earlier. Apart from the eye, such protected tissues include the testes (and we know Ebola virus can be maintained there for half a year) and brain and they are very generally described as 'immune privileged' sites.

The NEJM study also reported that the patient's semen contained RNA at 44 days after illness onset, and was still being monitored at publication. During his initial illness he had received the Ebola virus RNA-binding drug, TKM-100802 (or 'TKM-Ebola'[15]). It seems - for this one case anyway - that TKM-Ebola did not prevent persistence in at least two privileged sites.

This was not the first time an eye infection of this sort, called uveitis (inflammation of a region inside the eye), has been described in the EVD literature. In a study published about four years after an EVD outbreak in the Democratic Republic of the Congo, four ophthalmologist-confirmed cases were described. They developed symptoms between 42 and 72 days after onset of EVD, but no virology testing was reported.[9] In 1975, a report described a Marburg virus isolate grown from a fluid sample from the eye of a female with uveitis nearly three months after EVD onset.[10] All cases recovered their sight following treatment with atropine and steroid.

Apart from the NEJM story being a fascinating discovery and a tribute again to Emory's tenacity in trying to understand EVD, I have two concerns:

  1. According to the Disclosure forms attached to the NEJM article and signed by the authors, the article was likely submitted around February 2015 (amendment: Dr Varkey-lead author, confirmed via a Tweet to me, that the paper was submitted in January). It's now May 2015 and the EVD epidemic, while vastly reduced in scale compared to what it was in December, is ongoing. There have been growing reports of clinics devoted to dealing with the survivors of EVD and the description of post-Ebola syndrome is also growing.[8]
  2. The New York Times article included the following (bold is mine)...

    Dr. Jay Varkey, an infectious-disease specialist who had handled much of Dr. Crozier’s care, got special permission from the Food and Drug Administration to use an experimental antiviral drug taken in pill form. (The doctors declined to name it, preferring to save that information for future publication in a medical journal.)

    There is currently no evidence for or against any successful intervention using this unnamed drug. But the fact that the authors are considering publishing its use in a separate article could be interpreted as there being a chance it was helpful. Also, because the drug was obtained with special permission - it needed the drug company to agree to it being named. Perhaps they did not. But if this drug helped, it is a drug that should be made known, or available, to all who are seeing and treating EVD survivors. If the drug was of substantive use, then it might also be of helpful for survivors in Guinea, Liberia and Sierra Leone for slowing/eradicating virus in other immune privileged sites. I hope the company is considering this at least as urgently as vaccine companies did, before they eventually ramped up production of their products.
Publishing can often be slow 

The traditional process of publishing research results through scientific journals is a lot slower than many would like. This four month lag is an example of that. There are some journals that are faster than others and some that are slower. Publishing in some journals is traditionally considered to guarantee (you and) your data greater exposure. Some journals are where your data go to die. This is not just because of the journal per se, but also because of your data and because some journals are watched more closely than others by those who can make the complex science more understandable - the science communicators. They write the stories that draw attention to the big discoveries.

But peer-review is an important part of publishing quality science. Peer reviewing - finding expert reviewers who will review a manuscript in their own time and getting them to return their comments in a reasonable time frame - it is quite a challenge and it can be the cause a lot of delay to the process of manuscripts becoming published papers. It is also often argued that peer-review is a broken system, nonetheless it is (still) what we have now.[11,12,13]

Publishing should sometimes be fast

When there is an active or ongoing outbreak of infectious disease, knowledge is power; the power to perhaps help save lives, reduce the burden of the disease, halt transmission, find the source, change the diagnostics, manage infected patients differently, alter the message, use a new drug - generally the power do good things for people at risk, as fast as possible. 

On these occasions when time is so important to health, putting new data related to an outbreak through the scientific publication pipeline may be the least effective way to get important messages out in a timely manner. Before and after an outbreak - whatever - but during the outbreak - think differently. We've seen some great leaps in free genomic (viral sequence) data from Ebola virus variants - with the important detailed scientific papers following up later.[14] We still have a way to go to embrace not just a culture of openness, but one of speedy openness.[16] Not just for sequence data but for clinical data and for use of investigational drugs at times of crisis too.

All of the important aspects of the latest Ebola virus findings may have been passed quickly along to public health institutions that disseminate the information and act upon it if and as necessary. Nonetheless such behind-the-scenes information has remained behind-the-scenes. 

While eye surgery is unlikely to be very frequent in west Africa, a risk is now known for ophthalmologists who may operate on, or collect samples from, EVD survivors. We now know of at least two immune-privileged sites that can harbour infectious Ebola virus, many months after signs of disease have passed and the surviving patient has left care; the other site being the testes.[20,21] 

Could the central nervous system be yet another site? Could much of the remaining post-Ebola syndrome symptoms be related to virus replicating here causing headaches and nerve-related pain and tingling/burning?[8] Perhaps Emory medical researchers are working on this next.

Ebola virus continues to challenge us as we learn important lessons about managing its many and newly discovered impacts on the health of the host. This virus excels at revealing and exploiting our weaknesses. For something that is not alive, it has certainly played merry havoc in showing up our many failures to communicate.


  1. Weeks After His Recovery, Ebola Lurked in a Doctor’s Eye
  2. Persistence of Ebola Virus in Ocular Fluid during Convalescence
  3. Mystery ‘post-Ebola syndrome’ emerges in West Africa
  4. Sierra Leone: Helping the Ebola survivors turn the page
  5. Immune privilege or privileged immunity?
  6. The testis in immune privilege.
  7. Immune Privilege of the Testis: Meaning, Mechanisms, and Manifestations
  8. Mystery ‘post-Ebola syndrome’ emerges in West Africa
  9. Late Ophthalmologic Manifestations in Survivors of the 1995 Ebola Virus Epidemic in Kikwit, Democratic Republic of the Congo
  10. Outbreake of Marburg virus disease in Johannesburg
  11. Scientific Peer Review Is Broken. We’re Fighting to Fix It With Anonymity
  12. Is the Peer Review Process for Scientific Papers Broken?
  13. Is Peer Review Broken?
  14. Using Genomics to Follow the Path of Ebola
  15. Ebola: Tracking the Latest Measures Against a Killer
  16. Data sharing: Make outbreak research open access
  17. Status of U.S. Ebola cases
  18. Emory Healthcare Ebola Preparedness Protocols
  19. Emory's "Team Ebola" to receive National Patient Safety DAISY Award for exceptional nursing
  20. Ebola virus in semen is the real deal....
  21. Ebola virus in the semen of convalescent men

Wednesday, 6 May 2015

Snapdate: Confirmed Ebola virus disease cases - the end in sight?

I think we're a little bit beyond "jinxing" something by pointing it out, so here is graph of the confirmed Ebola virus disease cases based on the World Health Organization report date (Situation summary or Situation Report), including a basic model to predict when cases may hit zero, if nothing changes.

The P-value for this linear trend model is 0.00067. The standard error = 19.29;R-square = 0.14.
Click on graph to enlarge

I use Tableau Desktop Public Edition v9.0.0 for my graphs these days - and have taken advantage of its inbuilt linear trend model for this chart. This "model" accounts for precisely nothing apart from the trend based on the numbers that are available and have been plotted in this particular way, on this day, near a full moon. 

Reported numbers or outbreaks could flare up tomorrow or dry up overnight. 

I can say that over the past 2 weeks, data from each new summary or report have moved the predicted "end" data closer - from mid-June to now early June.

I am not an expert at modelling or statistics so please just take this at face value. The line suggests that if all things stay the same, we will reach zero considered cases per report around the 3rd of June 2015.

Please let it be so. 

Realistically, we may be heading for another "step down" - followed by a smaller trickle of ongoing cases for some period, ahead of a final push to zero. But there are experts who will know more about this than I.

Once we get to zero, the 42 day count begins.

Tuesday, 5 May 2015

The mechanics of the polymerase chain reaction (PCR)...a primer

The polymerase chain reaction (PCR) is a technique for copying a piece of DNA a billion-fold. As the name suggests, the process creates a chain of many pieces, in this case the pieces are nucleotides and the chain is a strand of DNA.

PCR is an enzyme-mediated reaction, and as with any enzyme, the reaction must occur at the enzyme's ideal operating temperature. The enzymes that are used for the PCR are DNA-dependent DNA polymerases (DDDP) derived from thermophilic (heat-loving) bacteria. As such, the enzymes function at higher temperatures than the enzymes we commonly use in the laboratory or have working in our bodies. These DNA polymerases operate at 60-75°C, and can even survive at temperatures above 90°C. This is important because a part of the PCR requires that the reaction reaches ~95°C as we shall see.

Apart from the DNA polymerase, PCR needs a DNA template to copy, and a pair of short DNA sequences called oligonucleotides or "primers" (described here) to get the DNA polymerase started.

Broadly speaking, there are 3 steps identified by incubating at different temperatures. The 3 steps make up a PCR "cycle".
  1. Double-stranded DNA separation or denaturation (D in Figure 1)
  2. Primer annealing to template DNA (A in Figure 1)
  3. Primer extension (E in Figure 1)

Figure 1.A PCR cycle.
The three temperatures which make up a single cycle. The DNA denaturation section (D), oligonucleotide annealing section (A) and the primer extension (E) section are marked. The temperature range over which dsDNA duplexes can denature (TD) or 'melt', and the range over which the oligonucleotide primer can hybridize (TM) are also marked.


At temperatures above 90°C, double-stranded DNA denatures or "melts". That means the weak hydrogen bonds that usually hold the two complementary strands together at normal temperatures are disrupted resulting in two single stranded DNA strands (shown below in an idealised form).

Primer Annealing..

At the annealing temperature (TA), primers that collide with their complementary sequence can hybrdise or "bind" to it. The chance of such an encounter happening is increased because we use a vast excess of each primer in the reaction mixture compared to the number of template molecules present.

The assay in the example below has been designed to amplify a region of the template spanned by and including, the primer sequences.

Primer Extension..

At the extension temperature (TE), the DNA polymerase binds to the hybridized primer and begins to add complementary nucleotides (i.e. every time the polymerase reads a "G" on the template strand, its adds a "C"; an "A" for a "T"; a "G" for a "C" and a "T" for an"A"), chemically binding each new addition to the last to form a growing chain. The process only occurs in one direction. In our example, the green primer is binding to its complementary template sequence and is facing toward the right (this is called the 5' (five-prime) to 3' (three prime) direction. Extension occurs in the direction that the primer faces. The result is a new double-stranded PCR product we usually call an "amplicon". An amplicon can be defined as an amplified molecule of a single type, in this case, an exact replicate of the original template.

Exponential Template Duplication..

The process is then repeated by cycling through the temperatures over and over again (35 to 55 times). Each cycle results in a new DNA duplex, each strand acting as a potential template for one or other primer.

Some interesting things stand out from the figure below.

The original template strands (blue and red) continue to act as templates because the PCR process is not destructive. However, each cycle produces a greater number of the shorter amplicon molecules. These are shorter in our example because the primers shown, bind within the template sequence. Eventually the majority of the amplicon in the reaction vessel will be the expected length, i.e. just the region spanned by, and including the primer sequences.

It is possible to mathematically predict the pattern of amplicon accumulation. In our example, we have started with two strands. In a perfect PCR reaction (which rarely occurs!), we have two new strands making a total of four. After the second cycle we have eight strands, then 16, 32 and so on. The reaction is doubling the number of strands each cycle or to make that an equation, we have 2n

Note: to make the process easier to understand, I have drawn the DNA strands as straight lines - in reality, DNA does not exist in as simple a form as this.

Further reading...

  1. PCR primers...a primer!
  2. Reverse transcription polymerase chain reaction (RT-PCR)...a primer
  3. Mackay IM. Real-time PCR in the microbiology laboratory. 2004. Clin Microbiol Infect. 10(3):190-212.
  4. Mackay IM, Arden KE and Nitsche A. 2002. Real-time PCR in virology. Nucleic Acids Res. 30;6. 1292-1305. 
  5. Beld MGHM, Birch C, Cane PA, Carman W, Claas ECJ, Clewley JP, Domingo J, Druce J, Escarmis C, Fouchier RAM, Foulongne V, Ison MG, Jennings LC, Kaltenboeck B, Kay ID, Kubista M, Landt O, Mackay IM, Mackay J, Niesters HGM, Nissen MD, Palladino S, Papadopoulous NG, Petrich A, Pfaffl MW, Rawlinson W, Reischl U, Saunders NA, Savolainen-Kopra C, Schoildgen O, Scott GM, Segondy M, Seibl R, Sloots TP, Wang Y-W, Tellier R and Woo PCYl. Chapter 10:"Experts’ roundtable: Real-time PCR and microbiology”, In: Real-Time PCR in Microbiology, IM Mackay (Editor). 2007. Caister Academic Press, Norfolk, UK.
  6. Mackay IM, Arden KE, Nissen MD and Sloots TP. Chapter 8. “Challenges facing real-time PCR characterisation of acute respiratory tract infections”, In: Real-Time PCR in Microbiology, Mackay IM (Editor). 2007. Caister Academic Press, Norfolk, UK. 269-317.
  7. Mackay IM, Mackay JF, Nissen MD and Sloots TP. Chapter 1: ”Real-time PCR; History and fluorogenic chemistries”, In: Real-Time PCR in Microbiology, IM Mackay (Editor) 2007. Caister Academic Press, Norfolk, UK.
  8. Mackay IM, Bustin S, Andrade JM, Kubista M and Sloots TP. Chapter 5:”Quantification of microorganisms: not human, not simple, not quick”, In: Real-Time PCR in Microbiology, IM Mackay (Editor). 2007. Caister Academic Press, Norfolk, UK.
  9. Mackay IM, Arden KE and Nitsche A. Real-time fluorescent PCR techniques to study microbial-host interactions. Methods in Microbiology, Microbial Imaging. (2005) Vol 34. Chapter 10.Elsevier. pp255-330.
  10. Mackay IM. Respiratory viruses and the PCR revolution. In: PCR Revolution: Basic technologies and applications, Bustin, SA (Editor). 2010. Ch 12. Pp189-211. Cambridge University Press.

PCR primers...a primer!

A DNA Down Under post
PCR (described here) functions mainly because of two components - a thermostable DNA polymerase and a pair of DNA 'primers'. Primers are short, made to order, stretches of oligonucleotides ('oligos' - from Greek meaning scanty or few). Modern oligos can be synthesized in lengths >100nt however the behaviour of oligonucleotides longer than 20nt is different from that of shorter oligos and different calculations are employed to determine their thermodynamic characteristics.

What is a primer...?

Primers, as their name may suggests, prime the nucleic acid template for the attachment of the polymerase. This is the first step towards duplicating that template. The primer directs the polymerase to move in a 5' to 3' direction (drawn left-to-right; Figure 1) because of the 'direction' of
DNA (See DNA Structure for more background).

Figure 1. DNA has direction. The polynucleotide chain shown 
above is 'read' in a 5' to 3' direction by the polymerase. This would 
be from the top to the bottom or from the phosphate group 
to the hydroxyl group.
Primer binding...

Primers hybridize at a temperature that is affected by their sequence, concentration, length and ionic environment. This annealing temperature is usually referred to as the TM (melting temperature) but is in fact 5–10°C below the TM. The term TM describes the temperature at which 50% of the primer–target duplexes have formed.

Primer specificity...

PCR gleans its extreme specificity from the primers. At each and every position of a new primer, we have 4 nucleotides to choose from, dATP, dCTP, dGTP and dTTP. 

Figure 2. Deoxynucleotide triphosphates. Each of the five deoxynucleotides are shown.2'-deoxycytidine-5'-triphosphate (dCTP; C9H16N3O13P3, MW=467),2'-deoxyguanosine-5'-triphosphate (dGTP; C10H16N5O13P3, MW=507),2'-deoxyadenosine-5'-triphosphate (dATP; C10H16N5O12P3, MW=491),2'-deoxythymidine-5'-triphosphate (dTTP; C10H17N2O14P3, MW=482) and 2'-deoxyuridine-5'-triphosphate (dUTP; C9H15N2O14P3, MW=468). 
So, if we design a sequence-specific primer of 20-30nt nucleotides in length ('20-30mer'), the chance that that exact sequence will occur randomly in nature will be 1/4 x 1/4 x 1/4 etc, 20 or 30 times i.e.

That means a 1 in 1012 to 1018 chance of a 100% homologous match to an unintended target. Or to put that in perspective, there are 2.85 x 109 basepaired nucleotides in the entire human genome.

While that all sounds very convincing, in reality, primers designed to detect viruses often share significant amounts of homology with the human genome - sometimes resulting in false positive amplifications. Even when the homology is far from 100%, primers may still amplify an unintended target as shown below. This most likely reflects the co-evolution of many viruses with humans during which time they have "captured" bits of our genome and "deposited" bits of their own genome. 

Next we'll list a few of the problems we can encounter when using the PCR.

Primer dimer...

The first problem I'll discuss is the most common and the most difficult to avoid. Depending on your requirements, it may also be the least significant.

When a small amplicon results from the extension of self-annealed primers, you get primer-dimer (PD) i.e. a dimer of one (self-annealing) or both primers resulting in a template capable of being extended by the polymerase. PD formation is highly efficient because the primers are in vast excess compared to the amount of template or even to the number of amplicon molecules at the end of the PCR.

This excess drives the formation of PD. Two main concerns arise from PD formation.

  1. Because PD formation is so efficient, it rapidly consumes dNTPs and primers and generates amplification inhibiting pyrophosphates. All of which can prematurely plateau the exponential accumulation of product.
  2. If we using a dsDNA-associating fluorescent molecule to follow the PCR's progress during real-time PCR, then PD will also show up, and, at least during the kinetic portion of the assay, cannot be differentiated from the signal of specific amplicon accumulation.
Figure 3. Examples of how primer-dimer (PD) amplicon can be formed.
Ten examples are shown of sense and antisense primer interactions
resulting in an amplicon. Note that different length amplicons can be formed.
The largest PD would result from the hybridization of the smallest number of
nucleotides and would approach the length of the two primers added together.

Mispriming is the result of a primer binding to an unintended template resulting in amplification. The amplicon (PCR product of a single species) can sometimes be the same size as the intended product, but is usually a different size when viewed following agarose gel electrophoresis.

Mispriming occurs because of poorly optimised conditions or because we haven't checked whether our sequence will inadvertently bind to an entirely different target entity e.g. a region of the human genome instead of the intended virus genome. Sometimes it just happens.

Mispriming can usually be avoided by more intensive comparison of the primer's sequence against the GenBank database using the Basic Local Alignment Search Tool (BLAST) at NCBI. Of course, a BLAST comparison will only find matches among those sequences housed in the database. When it comes to PCR where a single nucleotide mismatch can cause an amplification to fail, or at least perform with reduced efficiency, BLAST'ing primers can lead to a feeling of very false security.

In some instances the homology of the primer to its template may indicate a perfect match simply because viral variants have not yet been sequenced and submitted. Also, because there may be many undiscovered viruses and unsequenced non-viral genomes in the world, obviously none of which are represented on GenBank, a specific match, or a "no match", does not mean that you have exhausted your search for homologues. Take it all with a grain of salt. Designing two pairs of primers around the target region is a good place to start. This helps address the unexpected.

Structural problems...

This problem results from the way we design our primers. I am excluding self-annealing and secondary structures from here because we will deal with them specifically in the next section. --work in progress---

Further reading...
  1. A crowd-sourced database of virus primers.
  2. The mechanics of the polymerase chain reaction (PCR)...a primer
  3. Reverse transcription polymerase chain reaction (RT-PCR)...a primer
  4. Mackay IM. Real-time PCR in the microbiology laboratory. 2004. Clin Microbiol Infect. 10(3):190-212.
  5. Mackay IM, Arden KE and Nitsche A. 2002. Real-time PCR in virology. Nucleic Acids Res. 30;6. 1292-1305. 
  6. Beld MGHM, Birch C, Cane PA, Carman W, Claas ECJ, Clewley JP, Domingo J, Druce J, Escarmis C, Fouchier RAM, Foulongne V, Ison MG, Jennings LC, Kaltenboeck B, Kay ID, Kubista M, Landt O, Mackay IM, Mackay J, Niesters HGM, Nissen MD, Palladino S, Papadopoulous NG, Petrich A, Pfaffl MW, Rawlinson W, Reischl U, Saunders NA, Savolainen-Kopra C, Schoildgen O, Scott GM, Segondy M, Seibl R, Sloots TP, Wang Y-W, Tellier R and Woo PCYl. Chapter 10:"Experts’ roundtable: Real-time PCR and microbiology”, In: Real-Time PCR in Microbiology, IM Mackay (Editor). 2007. Caister Academic Press, Norfolk, UK.
  7. Mackay IM, Arden KE, Nissen MD and Sloots TP. Chapter 8. “Challenges facing real-time PCR characterisation of acute respiratory tract infections”, In: Real-Time PCR in Microbiology, Mackay IM (Editor). 2007. Caister Academic Press, Norfolk, UK. 269-317.
  8. Mackay IM, Mackay JF, Nissen MD and Sloots TP. Chapter 1: ”Real-time PCR; History and fluorogenic chemistries”, In: Real-Time PCR in Microbiology, IM Mackay (Editor) 2007. Caister Academic Press, Norfolk, UK.
  9. Mackay IM, Bustin S, Andrade JM, Kubista M and Sloots TP. Chapter 5:”Quantification of microorganisms: not human, not simple, not quick”, In: Real-Time PCR in Microbiology, IM Mackay (Editor). 2007. Caister Academic Press, Norfolk, UK.
  10. Mackay IM, Arden KE and Nitsche A. Real-time fluorescent PCR techniques to study microbial-host interactions. Methods in Microbiology, Microbial Imaging. (2005) Vol 34. Chapter 10.Elsevier. pp255-330.
  11. Mackay IM. Respiratory viruses and the PCR revolution. In: PCR Revolution: Basic technologies and applications, Bustin, SA (Editor). 2010. Ch 12. Pp189-211. Cambridge University Press.

Reverse transcription polymerase chain reaction (RT-PCR)...a primer

This post has been moved to the new Virology Down Under platform on Wordpress.

You can get to this specific post by clicking on the link below...

Please adjust your bookmarks.

Apologies for any inconvenience.

Monday, 4 May 2015

The third outbreak of influenza A(H7N9) virus seems to be over...

Cumulative curves of reported H7N9 cases and deaths in humans.
Click on graph to enlarge.
By the looks of the curve on the right, the rush of cases that defined the third known outbreak of the low pathogenicity avian influenza A virus subtype, H7N9, is over...for another season anyway. 

If we get into the nitty gritty, as I have below, there are a couple of interesting things to see. First though - let us remember that these are just reported data:

  • there may have been some cases that were not reported for whatever political, medical, social or personal reasons - these data are an idea of what happened - look at the rends and don't get hung up on the specific values
  • the overwhelming majority of the cases have reported to a healthcare facility with respiratory disease due (presumably) to either infection - we have no idea how many other people have been infected, what proportion were mild or asymptomatic (as I've discussed e.g. here and here, so we know it is possible). It could be half as many again, or 100 or 1,000 times as many.
  • these are only cases that have been examined with a laboratory test (as far as we know) - there may have been many other cases of "influenza-like illness" that did not get sampled and tested but were managed under (or not) the assumption that they were influenza of some type, subtype or strain.

Please note-the graphs used here can all be found on my fixed interactive H7N9 page at:

The interesting stuff includes:

  • For the 2 outbreaks we have continuous data for - 2013-14 and 2014-15, the start of the outbreak seems to be around October/November, with the peak around January/February. 
  • Outbreak 3 did not seem to reach the heights of the preceding year however, from what we could glean from pretty poor data, the link to poultry exposure was as strong as ever. Perhaps market closures in response to deaths were a little more effective/efficient/wide-ranging in Outbreak #3? Pure speculation

Click on graphs to enlarge.
  • Most of the cases in 2015 (bottom maps) were on the east coast of China

Click on map to enlarge
  • Most of the activity in the 3rd outbreak was focussed in Guangdong, Fujian and Zhejiang provinces (the red ones above) 
  • Xinjiang Uyghur Autonomous region (in the far west) and Guizhou and Hubei provinces joined the list of host regions in Outbreak #3
  • Xinjiang joined Guangxi and Jilin provinces (which reporting cases in Outbreak #2) as regions of China that share a border with another country - heralding the movement of this H7N9 variant beyond China's borders possibly into Vietnam, North Korea, or a -stan
Click on graphs to enlarge.
Keep an eye out for H7N9 Outbreak #4 - coming to a colder China around November 2015. 

But for now, it might be time to hit the FluTrackers line lists (okay, I've had 4 tabs open for ages) and graph the course of another source of concern - H5N1 cases in humans.

Hubei province listed its first H7N9 case in April...some rare detail

A new province was recently added to the list of those reporting cases of avian influenza A(H7N9) virus infection in humans. 

Of course reporting does not mean capturing. Reporting has been weak this season. The cases that have shown seem to be just those who were ill enough to visit a Doctor/hospital and get a laboratory test. This is the same story for most infectious agents. We see just the tip of the iceberg, the beginning, the head of the arrow, we only scratch the surface, the glycoprotein on the envelope of the virus as it leaves the...okay, you get the picture.

This year has seen a very disappointing effort by China to provide useful public data that could permit tracking of what has become the annual outbreak of human cases of avian influenza A(H7N9) virus infection.

The human H7N9 case hotzone, at least since we heard about the virus infecting a human in February of 2013, have been on the east coast of China. We currently stand at 659 reported human cases, and over 200 deaths. Very. Roughly.

Click on image to enlarge.

I remember fondly a time when there were scads of data on H7N9-related human cases and deaths. Okay, China did over-share on a number of occasions....

Click on image to enlarge.
Story to be found here.
Click on image to enlarge. 
Story to be found here.
Click on image to enlarge. 
Story to be found here.

..but things have changed. 

For a comparison take the Kingdom of Saudi Arabia's Ministry of Health and their efforts to provide public data on Middle East respiratory syndrome (MERS) and its coronavirus (MERS-CoV). While there are a few gaping holes in the data set (c'mon guys-fill these in!), there can be as many typos as on this blog (but I'm not a public health Ministry - in case you were wondering) and the data can be intermittent, it represents the best public source of detailed, yet deidentified, human data on an ongoing zoonotic viral emergence. And that's saying something. But congratulations nonetheless!

I'm not including Ebola virus disease data-gathering here - the fact that we have had so much data - despite the initial lack of infrastructure and people trained to collect, collate and report that detail - is a fantastic testament to the efforts of those on the ground in Guinea, Sierra Leone and Liberia.

But this season H7N9 data that have been reported by public health sources have been released in blocks and lack any consistent or useful detail, except the province. Some detail is available when harvested from media reports by the ever assembled FluTrackers team. I rely heavily on their line list (to be found here). 

One example of the poor data quality this season, take a look at this text from a recent Disease Outbreak News provided by BigBlue (that's the World Health Organization, or WHO, for those not accustomed to my street groove)... 

No-one will be reading that and feel overly informed.

One is left to assume that this is how these data are coming out of China - infrequently and without detail. We regularly see that when better data are provided - and again, I hold up MERS-CoV case descriptions here - they get publicly listed by BigBlue. 

And before you head to your keyboard to ask "Why should we have access to these data?"...I will first ask you - why shouldn't we? They are collected and collated internally. They are of interest to epidemiologists, model builders, public health planners and data tinkerers the world over. And it's not as though the details are subsequently released in peer-reviewed publications. They are not. 

It's just disappointing.