Category Archives: Software

Hygiene hypothesis should be death but is still alive

There is a new paper today “The Hygiene Hypothesis in the Age of the Microbiome“. Fortunately there is a strong paywall…

An impressive illustration of the hygiene hypothesis was found in the consistent farm effect on asthma and allergies, which has partly been attributed to immunomodulatory properties of endotoxin as emitted by livestock.

which is not aware of “Time to abandon the hygiene hypothesis: new perspectives on allergic disease, the human microbiome, infectious disease prevention and the role of targeted hygiene

Although evidence supports the concept of immune regulation driven by microbe–host interactions, the term ‘hygiene hypothesis’ is a misleading misnomer.

see also “News Feature: Cleaning up the hygiene hypothesis

Still, the catchy hygiene hypothesis continues to be widely embraced by the public, the media, and even scientists: Uses of the term in the scientific literature rose threefold over the past 10 years compared with the decade prior, according to a search on Thomson Reuters Web of Science. “In science, when something has been propagated for so long, it can be hard to change,” says Marsha Wills-Karp, chair of environmental health and engineering at Johns Hopkins Bloomberg School of Public Health.


CC-BY-NC

Open letter to the Sente developer

Dear Michael J. Cinkosky,

this is an open letter regarding Sente academic library software.

Sente is used by tens of thousands of researchers in all academic fields at universities and research organizations around the world

I am using Sente nearly every day for many years to collect and annotate the literature.

It is useful to retrieve papers and insert references in my own papers.

It seems that Sente is not more supported.

A while ago, I could build up a bibliography only in exports from my text processor.

Then online sync stopped.

Most recently also Google Scholar reference recognition ended.

It is now a useless software. I paid for it, I used it, which was a fair deal.

But now am now lost, if Sente just drowns in the internet.

Could you sell it to a company that continues support?

Your customer base would really appreciate it.

Or write a final countdown Endnote export script that preserves PDF links.

Or just give us the code at Github, we will continue to develop it.

We all appreciate your work with Genbank and also with Sente.

But don’t leave us without saying good-bye.

All the best,


CC-BY-NC

Ende der Debatte

aus dem DH Newsletter 11/2017

DHV sieht Debattenkultur in Gefahr
Der Präsident des Deutschen Hochschulverbandes (DHV), Professor Bernhard Kempen, hat vor einer Erosion der Debattenkultur an Universitäten gewarnt. “Das Klima der Political Correctness ist bedenklich”, führte er im Interview mit der “Welt” aus. “Die Sensibilitätsschwelle für andere Ansichten, andere Meinungen, für das Fremde ist erheblich gesunken.” Früher seien Universitäten Stätten vehementer und teils heftiger geistiger Auseinandersetzungen gewesen, heute werde versucht, niemandem eine Ansicht zuzumuten, die ihn verstören könnte.


CC-BY-NC

Sommerschlussverkauf

academics.de zu Hochschule 4.0 online Nov 2017

Harvard und MIT gründeten 2012 gemeinsam die Plattform EdX, auf der hochrangige Universitäten aus der ganzen Welt MOOCs anbieten. Auf der Plattform udacity werden “nanodegrees” (Mini-Abschlüsse) angeboten, die von Unternehmen mitkonzipiert und -finanziert werden (z.B. “Ingenieur für künstliche Intelligenz” in Zusammenarbeit mit IBM Watson). Wie “Hochschule 4.0” und das Forschen, Lernen und Lehren letztendlich aussehen werden, in welchen Kurs- und Lernformaten und mit welchen Abschlüssen die Vermittlung von Wissen zukünftig gestaltet werden wird, ist eine offene Frage.

detailreicher, besser geschrieben, aber auch kritischer: “Kein Bock auf MOOC” Dirk Böttcher und Lea-Marie Kanzler.
Links: https://www.class-central.com https://www.edx.org https://de.udacity.com


CC-BY-NC

Correlation not causation

I am following up some references Rudi Balling showed in his talk yesterday. Sure, correlation is not causation,

An excerpt of the famous Wright paper 1921

but I fear that with all that systems biology we already ended in the scientific nirwana than generating any useful hypothesis.


CC-BY-NC

Vertrauensverlust

Forschung und Lehre heute

Zwölf Prozent der Deutschen haben kein Vertrauen in Wissenschaft und Forschung, jeder zweite bekundet grundsätzlich Vertrauen. Der Rest (37 Prozent) zeigt sich unentschieden. Vor allem die Expertise der Forschenden wird als Grund für Vertrauen aufgeführt. Als häufigster Grund für Misstrauen wird die Abhängigkeit von Geldgebern genannt. Das geht aus dem Wissenschaftsbarometer 2017 hervor. Die bevölkerungsrepräsentative Umfrage betrachtet seit 2014 jährlich die Einstellungen der Öffentlichkeit gegenüber Wissenschaft und Forschung. Sie wird von Wissenschaft im Dialog (WiD) herausgegeben und von der Robert Bosch Stiftung gefördert.

Barometer 2017


CC-BY-NC

Universities are being hollowed out

This letter was just published in the Irish Times

The parlous impact of managerialism on the sector is only too obvious: 52.4 per cent of all staff in Irish universities are not academics yet academics have seen a huge rise in their administrative duties to the detriment of time for research, teaching and contact with their students; the huge salaries commanded by presidents of universities, who see themselves as captains of industry, contrast ignominiously with the relative pittance earned by the rank-and-file lecture staff; and the cynical reduction of the number of tenured staff has resulted in 45 per cent of academics being employed part-time on salaries as low as perhaps €30 an hour, with no job security and no facilities to develop their teaching and research (many do not even have borrowing rights for the university library, a place to hang their coat, or an allowance to cover marking).


CC-BY-NC

Live votes during presentation

Live votes can increase participation of your audience but there are quite some technical challenges.

For my most recent presentation I have set up my own Wifi using an old Fritz! Box flashed with Freeze that can route all incoming traffic towards a Macbook (that is connected to the router by cable).

The Macbook itself is running a webserver that delivers the questions to the mobile phones in the audience. Mobile clients basically see this page

 

The presentation with all results is running in a Chrome full screen windows that is mirrored to the projector. It is basically an export from Powerpoint pictures that are shuffled around by a javascript framework (bespoke.js). Results are stored in an adhoc sqlite database and charts are produced on the fly by d3.js.

All scripts are available per email request.


CC-BY-NC

Datafication of Health

I had the unique chance yesterday to hear a talk of Minna Ruckenstein (University of Helsinki) about datafication of health. I think this is one of the most important topics of the last decade – being one of the largely underestimated, largely ignored development, but nevertheless relevant for the future of medicine. It basically means conversion of health into data (like Facebook is converting communication into data) which happens on multiple scales and at multiple levels. Her excellent paper at Annual Review of Anthropology was just posted a month ago and can be downloaded at the journal website (or at SciHub).

Minna Ruckenstein, ChipMe Meeting, Galway, Ireland

Recognizing that the datafication of the traditional health ecosystem is generating new power asymmetries and disrupting traditional regulatory and ethical mechanisms, some scholars have embarked on applied research projects, often collaborating outside their academic spheres. These citizen-centered initiatives appeal to patient organizations, savvy self-trackers, and ordinary people to address issues such as the tension between data openness and data ownership; asymmetries of data usage and distribution; the inadequacy of current informed consent and privacy protections; and the need to reappropriate and rearticulate concepts such as “sharing” and “the public good” that have been co-opted by technology companies seeking free access to their users’ data.


CC-BY-NC

Bitcoin is a bad idea

Nature today thinks “It’s good to have lots of bad ideas”. Bitcoins are a bad idea as they are traded in for energy consumption.

There are lots of different bitcoin mining computers out there, but in recent months, companies have focused on ASIC miners, which use less energy to conduct their calculations. Mining companies that run lots of ASIC miners as businesses have told me that they use one watt of power for every Gigahash per second of computing that they do when mining for bitcoins.
At this rate, the bitcoin network runs at 342934450 watts, which equates to around 343 megawatts. Calculations based on EIA data reveal that the average US household consumes about 1.2 kilowatts of power, meaning that 343 megawatts would be enough to power 285,833 US homes at the time of writing (May 2015).
That’s quite a lot of energy – about a third of the homes in San Jose. And our 1 watt per Gigahash/second figure is pretty efficient. There will be a large number of residential users who will be taking more power to run their miners.So it’s likely that this is a conservative estimate.

Another estimate by digiconomist.net is 16 TWh per year (Aug 2017). Kilo 10^3, Mega 10^6, Tera 10^9…


CC-BY-NC

Goldwater Rule

The Goldwater Rule makes sense.

The Goldwater rule is the informal name given to Section 7 in the American Psychiatric Association’s (APA) Principles of Medical Ethics,[1] which states it is unethical for psychiatrists to give a professional opinion about public figures they have not examined in person

But whenever an individual releases more answers than you want to ask him questions, a preliminary diagnosis can be made. So think Sword and Zimbardo

As Trump’s campaign, and his narcissism, gained momentum, so did our efforts. In March 2016, we posted The Narcissistic Personality: How to Spot a Narcissist, in which we shared clinically documented narcissistic behaviors and hoped it would be easy for readers to see that Trump fit every example.

So thinks Lance Dodes

Mr. Trump’s speech and actions demonstrate an inability to tolerate views different from his own, leading to rage reactions. His words and behavior suggest a profound inability to empathize. Individuals with these traits distort reality to suit their psychological state, attacking facts and those who convey them (journalists, scientists).

And so thinks also Robert Lifton along with many others.


CC-BY-NC

Es sei außerordentlich sinnvoll

Michael Ristow – bekannt durch eigens verfasstem Wikipedia Eintrag zu „Mitohormesis“ aber keinem einzigen Vitamin D Artikel – äussert sich zur Vitamin D Versorgung von Kleinkindern im Deutschlandfunk

Es sei außerordentlich sinnvoll, dass sie regelmäßig diese Präparate bekommen, sagte der Professor für Energiestoffwechsel an der ETH Zürich Deutschlandfunk. Zuvor hatten kanadische Mediziner eine klinische Studie mit über 700 Kindern im Alter zwischen einem und fünf Jahren durchgeführt. Die eine Hälfte der Kinder bekam in den Wintermonaten eine Standard-Dosis Vitamin D verabreicht – die andere Hälfte der Kinder erhielt eine besonders hohe Dosis Vitamin D. Das Ergebnis der im Fachblatt “JAMA” veröffentlichten Studie: Viel Vitamin D hilft nicht. Zwischen beiden Gruppen gab es demnach keinen statistisch signifikanten Unterschied in Bezug auf die Zahl der Atemwegsinfektionen. Internist Ristow betonte, die Forscher hätten nicht verglichen, wie Kinder ohne zusätzliche Vitamin-D-Versorgung im Vergleich zu jenem mit Supplementierung abschnitten. Das bedeute nicht, dass Vitamin D nicht Einfluss habe auf Erkältungshäufigkeit und -dauer haben könne.

Wie kann man den Sachverhalt nur so verdrehen?
Die JAMA Studie schreibt doch völlig korrekt in der Einleitung

Although a number of clinical trials have attempted to assess the effect of vitamin D supplementation on the prevention of respiratory tract infections among adults and children, conclusions have been hampered by small sample sizes, short trial duration, and lack of laboratory-confirmed outcomes. Likewise, a meta-analysis of 7 clinical trials found no association between vitamin D and childhood acute respiratory infections, but limitations included a shortage of studies and heterogeneity in terms of populations and end points. To our knowledge, only 1 trial has assessed the effect of vitamin D on laboratory-confirmed upper respiratory tract infections in children specifically. It included 334 children in Japan and found a significant reduction in influenza A, but not influenza B, among children receiving 1200 IU/d of vitamin D supplementation. It remains unclear if vitamin D supplementation can prevent all-cause upper respiratory tract infections among children.

Es gibt also keine wissenschaftliche Begründung, dass Kinder regelmäßig und anlasslos Vitamin D bekommen.


CC-BY-NC

99.9% of scientists are being forgotten after 20 years

The estimate in the last LJ magazine 07/2017 (Stephan Feller, “Jenseits des Hamsterrads”, p6) was probably too optimistic giving a figure of 90%. But when looking at some scholarly papers my estimate looks more accurate.

Citation amnesia has been described some 6 years ago as “A Systematic Examination of the Citation of Prior Research in Reports of Randomized, Controlled Trials”

In reports of RCTs published over 4 decades, fewer than 25% of preceding trials were cited, comprising fewer than 25% of the participants enrolled in all relevant prior trials. A median of 2 trials was cited, regardless of the number of prior trials that had been conducted. Research is needed to explore the explanations for and consequences of this phenomenon. Potential implica-tions include ethically unjustifiable trials, wasted resources, incorrect conclusions, and unnecessary risks for trial participants.

A more recent paper on “Attention decay in science” showed

this decay can be described by an exponential or a power law behavior, as in ultradi usive processes, with exponential tting better than power law for the majority of cases. The decay is also becoming faster over the years, signaling that nowadays papers are forgotten more quickly.


CC-BY-NC

Andreas Weigend, Zhang Zheng, George Church

Das “Who is Who” der Datensammler.

Andreas Weigend ist der frühere CSO vom Amazon. Interessanterweise verschweigt er das auf seiner eigenen Homepage und wird -danach gefragt- schmallippig im Interview. Ein Physiker mit Interesse an neuronalen Netzen, Deutscher mit Wohnsitz San Francisco und Shanghai.

Zhang Zheng, Peking, ist Ökonom und Parteisekretär. Bekannt geworden im Westen für BDA, Big Data Abuse. Jedem Einwohner wird ein Score zugeordnet wird, der letztlich über seine Rechte im Alltagsleben und  seinem Zugang zu gesellschaftlichen Ressourcen entscheidet.

George Church, US Amerikaner, Genetiker in Boston, mit einer Chinesin verheiratet, bekannt vor allem durch das Personal Genome Project, bei dem man seine komplette Genomsequenz abliefern soll. Zehn Jahre nach Gründung und nur 5500 von geplant 100.000 Freiwilligen, ist zumindest der europäische Ableger von PGP seit längerem auf Tauchstation.

Drei Proponenten von Big Data also, die ihre ehrenwerte Motivation immer wieder beteuern: das Einkaufserlebnis zu steigern, das soziale Miteinander optimieren und die Gesundheitsversorgung zu verbessern. Kollektivistisches Gedankengut, nicht umsonst taucht in allen drei Bespielen China auf.

Das deutsche Bundesverfassungsgericht erklärte die deutschen Vorschriften zur Vorratsdatenspeicherung mit Urteil vom 2. März 2010 für verfassungswidrig und nichtig. Das Urteil verpflichtete deutsche Telekommunikationsanbieter zur sofortigen Löschung der bis dahin gesammelten Daten. Zur Begründung gab das Gericht an, dass das Gesetz zur anlasslosen Speicherung umfangreicher Daten sämtlicher Nutzer elektronischer Kommunikationsdienste keine konkreten Maßnahmen zur Datensicherheit vorsehe und zudem die Hürden für staatliche Zugriffe auf die Daten zu niedrig seien

Und natürlich stimmt es nicht, daß hier das Einkaufserlebnis verbessert wird. Wenn, Beispiel Weigend, Amazon mir den Kauf verweigert weil ich schon zuviel umgetauscht habe, dann optimiert Amazon lediglich seinen Ertrag,  Nicht besonders ehrenwert bei einem Konzern, der hier keine Steuern zahlen will. Das mit dem sozialen Miteinander ist eine Paraphrase für faschistische Allmachtsphantasien. Und wer auf bessere Gesundheitsvorsorge hofft, liest am besten bei Sascha Lobo nach “Warum bezahle ich Rentenversicherung, wenn der Computer ohnehin sagt, dass ich mit 50 sterbe?”

Mal abgesehen davon, dass in allen Datensätzen massive Fehler stecken, dass Algorithmen an manchen Kombinationen scheitern, zu kurz in der Diskussion kam vor allem eines: Big Data Algorithmen sind massiv gegen den demokratischen Rechtsstaat gerichtet. Es gibt keine Instanz der Beschwerdeführung, es gibt keine unabhängige Gutachten, es gibt keinen Verteidiger und keinen Richter.

Weigends neurales Netz hat dich bei Amazon gesperrt, basta.
Zhengs App “Ehrliches Shanghai” entscheidet Dich unter 849 Punkte zu setzen, basta.
Churchs Sequenzscore gibt Dir nur 50 Lebensjahre, basta.

Bei Amazon muss man nicht einkaufen, in China muss man nicht leben, und Gensequenzen kann man ignorieren.


CC-BY-NC

Zero probability

Thomson Reuters Citation Laureates is a ridiculous list of candidates considered likely to win the Nobel Prize in their respective field. The candidates are so named based on the citation impact of their published research. Zero probability for the last 3 years.


CC-BY-NC