Wednesday, April 30, 2008

Powerpoint presentation tips

by Beyond Lab

Note: one slide was a picture taken on one of the annual national big decision-making meetings in China showing the audience was bored and felt asleep. This is not a result of bad powerpoint presentation though, -- the presenter (one of the leaders) was simply reading what everyone has in his/her hands.

Another funny one:

Want more:

How not to give a presentation: Richard Smith, editor; BMJ

Tuesday, April 29, 2008

on scientific language

by Beyond Lab

T. Ryan Gregory (an assistant professor of evolutionary biology and genome biology at the University of Guelph in Guelph, Ontario, Canada) tried to translate scientific expressions in papers to plain language in a humorous way. This is fine. But please don't take it seriously because it is absolutely not always true.

Maybe evolution people do that (for example to hand-draw a regression line), but not all other fields can do that. Especially, if you do microarray analysis where you have millions of data points, you are always asked to deposit your raw data to a public database at NCBI, which everyone can access and analyze. --If there's any manipulation, some body will find out. Then you will be doomed.

Plus, most scientific conclusions must be reproduced by different labs before being generally accepted.

However, still he has a good point that the scientific language in published papers should be more meaningful to lay people. --Maybe a short plain version of every paper should be included while being published.

Friday, April 25, 2008

personal genome discussion

by Beyond Lab

A discussion of the personal genome by a panel of "masters" from various fields. Although so far no video available, but this is a must see.

At Univ. of Washington, on Apr 23, Bill Gates, Eric Lander, Maynard Olson, Leena Peltonen, and George Church discussed many aspects of the personal genome: the science, the industry, and ethics etc.

Some interesting notes from above Sandra Porter's post.

"The panelists were asked if they had had their genomes sequenced. Only George Church answered yes. It was interesting though, even though Leena Peltonen and Eric Lander said that they weren't interested in having this done, both of them said they had been tested for certain genetic diseases, Eric, for Tay Sachs, and Leena for 40 diseases common in Finns. Bill Gates said that if the top 20 infectious diseases were to be cured, he would be happy to have his genome sequenced and make it public."

"Eric Lander playfully asked the other panelists if they thought presidential candidates disclose their genomes. He reminded us that we had a president with Alzheimer's disease and we would have found that potential if we had tested him. In the future, will we ask the older candidates to get tested for Alzheimer's?"

--I heard Eric Lander's talks. It is always very informative and amusing. :)

Here you are. I did a quick google and found a place to watch the discussion.
The beginning part is just music --weird. but drag to start from after 15min.


Thursday, April 24, 2008

GINA, proteome, and more

by Beyond Lab

GINA will very possibly be passed today and sent to George Bush. This is good in protecting your gene (personal genome) information from being misused by insurance companies etc. To know about GINA, this blog has the details.

Personal proteome? It might be too early but it eventually will come around. Researchers around the world are trying to start an ambitious project to map all the proteins in our body. It will definitely cost more than sequencing our genome because there are so many types of cells/tissues in our body and each type has characteristics proteins. Moreover, to understand protein is not just know their names and sequences -- the higher order structures (2,3, and 4-dimensioin) and the interactions among proteins are all critical. How should this be funded? Public money? Private support? Scientists would have to do some marketing to get enough support. Nonetheless, it is useful and must be done.

A step further, once this information is available some day, will there be many private companies to sell personal proteome service like the current personal genome companies? Let's see.

On another topic, since most of the genome or proteome research is done with public funding, should the information be used by private companies to earn money??? Should they be charged for using that information for profit. --Interesting to think about it. Your comments are welcome.

Wednesday, April 23, 2008

more on personal genomics

by Beyond Lab

Just something to back up what I wrote in yesterday's post about personal genomics.
Boston Globe has an article about this
Some doubt genome's value as health tool.

There have been a lot of this kind of reports some where, both in scientific journals and media. However the image is still not clear. It would be interesting to see how the companies promote their products and challenge the scientist's views. Will simple search engine optimization do the work??

Tuesday, April 22, 2008

It's Google, again!

by Beyond Lab

Just wondering, do you want to have your personal genome information go public like James Watson? Be careful if you pick 23andme or Navigenics to have your DNA sequenced. Google just invested a big lot of money to Navigenics. Eventually, Google will probably put everything up to the web including indexing your DNA.

Just kidding. What I don't really understand is to me (and certainly a lot of other scientists in the field), personal genome or consumer genomics is certainly not ready yet, but why so many people are jumping into this water??!! Is it fun?

Maybe Google will put ad links to every possible website and to promote eMarketing of this 'promising' new thing. One day, conversation will start like this, "Sequenced yet?"

Monday, April 21, 2008

consumer genomics, educational or medical

by Beyond Lab

New York State sent letters to companies offering consumer genomics tests (personal genomome) because the state considers these tests as "medical" test and should obtain a permit before being prescribed by physicians. The 23 companies include Navigenics, 23andMe, Affymetrix, Illumina, and HairDx etc. Some companies claim consumer genomics is educational, therefore should not require special permit. What on earth is this?

--This is a gray area. But considering the massive information (potentially informative at many aspects), I tend to think that consumer genomics belongs to those genetics tests which are not simply educational. Proper regulations are necessary. FDA probably should hold some kind of panel discussion before it really becomes a problem. Cold Spring Harbor Laboratory is holding such a symposium late this year on this topic (both technology and applications). It is a quickly expanding area involving a lot of deciplines, which makes this kind of discussion important.

I am exciting to see what other people say. So please leave your opinions here.

Sunday, April 20, 2008

personal genome, internet and eMarketing

by Beyond Lab

I did a simple experiment today --search the internet for "personal genome" using different search engines --just want to see what you get on the first page. Here are the results. Only 23andme, which is literally part of google, shows up in google search. Seems some of the companies need to do some search engine optimization (SEO) if they want to have better internet marketing.



MSN Live:

PS. There are too many terms with basically same meaning. For example, if you try "consumer genomcis" on these search engines, you will get different hits. But the fact is NO companies were on the first pages.

Friday, April 18, 2008

alcoholism connected to epigenetics

Epigenetics gets real!!!

A group from Chicago published a paper two weeks ago connecting chromatin remodeling (epigenetics) in brain to alcoholism. It is exciting because this is for the first clear evidence, as far as I know, humans' everyday activity could affect epigenetics.

Using laboratory rats as model, they found that acute anti-anxiety effects of alcohol are mediated by reduced histone (H3 and H4) acetylation in the amygdala region of the brain. What's more interesting is that increased anxiety accompanied with withdrawal from alcohol dependence is caused by a further increase in this process (acetylation). Neuropeptide Y, a known protein and mediator of anti-anxiety, is a potential target of the transcriptional regulation associated with this epigenetic mechanism. Inhibition of this process could indeed prevent anxiety. These data could potentially explain a lot of symptoms related to alcoholism and withdrawal anxiety. The clinical significance will have to be proved by further study.

When you drink next time, remind yourself that you are modifying your epigenome!

Thursday, April 17, 2008

cancer clinical research and Sydney Brenner

Nobel laureate Sydney Brenner called for a “bedside to bench” approach to researchers at the AACR meeting this month. Biological and medical researches have been following a way from lab bench to hospital bed all the time. Input from clinical professionals to basic research is limited. This actually could be one of the most important reasons that translational research is extremely slow.

However, it is easier to say it than do it. This could be a slow process with both parts having to be patient enough. It is like talk to evolution people about molecular biology - macro world and micro world have different languages.

But this should benefit us and should be promoted. Go for it!

Wednesday, April 16, 2008

Watson and Alzheimer’s disease

You must have heard that the "complete" sequencing of the genome of James D. Watson - the pioneer of molecular biology. The sequencing actually was finished about one year ago in two months and cost about 1 million. Now the results are out in Nature.

Here are some numbers:
7.4 - fold redundancy -- current sequencing technique requires redundant sequencing for assembly purposes;
3.3 million - single nucleotide polymorphisms (SNP) --these are single nucleotide variations to the reference sequences;
0.61 million - of those SNPs were previously unknown;
10,654 -- 10654 of those SNPs could cause amino-acid substitution within the coding sequence, that means that the protein sequences could be changed which could affect the protein function;
222,718 - small insertion and/or deletion polymorphisms, which affects local chromosomal regions;
345 - of above overlap gene coding sequences and could alter protein function;
1.5 million -new sequences;
49 - potential genes from above 1.5 million new sequences

Will Dr. Watson develop Alzheimer's disease? That's secret. The sequence of Apoliprotein E gene and neighboring regions was not disclosed.

How about breast cancer? --come on, not him. But his genome does contain a related mutation. He is not too concerned because he doesn't have any daughter.

Are you interested in his genome? (I am not :) ) But here is the link:

Is Dr. Watson the first one to have genome completely sequenced? -- No. The founder of J. Craig Vender Institute (Rockville, Maryland), Mr. Vender was the first one. However, the technology was different and it cost $100 million. 100 times more!

How useful the sequence is? --"it will be extremely difficult to extract medically, or even biologically, reliable inference from individual sequences" - Maynard Olson.

How useful will it be? -- Some day, it will be useful.

Tuesday, April 15, 2008

stem cell patent?

Steven Salzberg, a Professor of bioinformaticsat at the University of Maryland, raised the question of patenting stem cells by WARF - Wisconsin Alumni Research Foundation. Many aspects involved here, for example, is patent a good thing at all? Is patenting research a good thing? Is patenting research funded by public money a good thing? Look at people's comments. Interesting topic to discuss.

My point is simple: patent them but give them free (or almost free) to academic researches.

Saturday, April 12, 2008

Stem Cell meeting

The 6th ISSCR (International Society for Stem Cell Research) Annual Meeting will be held from June 11 to 14, 2008 in Pennsylvania.

Looking at the speaker list, you will regret if you are in this field but you will miss this conference.

Conformed speakers include:

Sir John Gurdon FRS, Wellcome Trust/Cancer Research UK Gurdon Institute, UK, Keynote Address
Rudolf Jaenisch, MD, Whitehead Institute for Biomedical Research, USA, Closing Address
Douglas A. Melton, PhD, Harvard University, USA
Shinya Yamanaka, MD, PhD, Kyoto University, Japan

and more...

Friday, April 11, 2008

Parkinson's disease - new information

Parkinson's disease has been thought to be a good disease model where cell replacement could actually help. It's caused by the damage of a particular kind of neural cells by an abnormally formated protein. Therefore anything replacing these damaged cells are thought to be effective. Stem cell researchers around the world are trying to make this kind of cells from embryonic stem cells for replacement. However it should be noted that earlier trials in 1990s already started injecting normal cells from aborted fetuses into the affected patients' brains to replace those cells destroyed. Although not exactly same as stem cells, these researches did offer some useful information. First, the injected cells did actually survive and incorporated to the patients' brain system. Second, these cells could offer some benefit (although not dramatic in most cases).

Now, some of the patients have died and postmortem examinations told us something surprising -- in some cases, some of the injected cells acquired similar damage as the original disease cells. Parkinson's disease usually only hits old people. However those injected cells are still relatively "young" (slightly over one decade), how did these cells got the dangerous protein? No one could answer yet. Although only 6 cases were analyzed, these findings indicate the complex of this disease. Will the same thing happen to stem cell derived cells in the future? How to get away with this problem?

Further research is necessary. problems will be solved eventually.

Thursday, April 10, 2008

What can be done will eventually be done

-- in time!

just put it up here.

another step toward regenerative medicine -- clonal precursor cells

A recently online paper from Regenerative Medicine (May 2008, Vol. 3, No. 3, Pages 281-302) entitled The ACTCellerate initiative: large-scale combinatorial cloning of novel human embryonic stem cell derivatives might offer something toward both basic research and clinical medicine.

Michael D. West (BioTime, Inc, Alameda, CA) et al. used a shotgun-like, or random induction/selection, approach to isolate 140 cell clones (hEP as they named them)from human embryonic stem (ES)cells. These cells are not ES cells anymore. They are differentiated toward various different lineages but not terminally differentiated either. What are they? Not sure yet as of now because further characterization and comparison to all known cell types in our body is required. Are they useful? Yes. 1) these cells are relibable sources because ES cells are extremely sensitive to all kinds of environment factors and too difficult to control in experiment. Therefore ES cells with the same name from different labs may actually be different things. These hEP clones, once properly characterized, could certainly provide a reliable source for research. 2) From these, further study may induce them into different cell types needed in medical research/practice. Currently, people can only induce ES cells into a limited number of final differentiated cells and they are not pure at all. These certainly limited the progress of research and regenerative medicine. They didn't say this --but I think it is very interesting and practical -- following the same shotgun approach, these hEP cells could be induced to further differentiate down the road to become more specified cell types (randomly). Certainly, some of them should be actual tissue cells and could be useful. These might actually bypass the difficult targeted induction approach.

Ethical problems? Of course. For example, what exactly are those hEP cells or the further differentiated cells as I proposed? New entity?? Questions like these certainly exist but the research should move on because humans will benefit eventually. A quote from a pioneer Sir John. Gurdon, "if something works well, ethical concerns will disappear."

Wednesday, April 9, 2008

MethodShare from Genome Technology

Genome Technology Online has launched another method/technology forum called MethodShare. ( or

There are a few articles and some discussion about various life science techniques. But so far, I don't see anything really interesting or different from other general forums such as protocol-online. It is also not necessarily good if you are particularly interested in something for example python in bioinformatics -- you'd better go to more focused places.

Anyway, it is certainly a good e-Marketing strategy. And keep an eye on it!

Tuesday, April 8, 2008

not necessarily new but useful tips

These tips might help you increase the yield and quality of your DNA sample from gel extraction.

Although the manufactures don't really put these in their manual, actual bench scientist always have good tips for improving the experiments. --Talk to them or visit these blogs. --Will help.

Monday, April 7, 2008

Blogging your thesis?!

This guy, a PhD candidate in microbiology and currently at Tulane University, is blogging his ongoing thesis writing online.

With his advisor's approval and several journal editors' positive feedback, he's posting part of his thesis, namely the introduction (which is a kind of review), and probably material and methods. As for the unpublished result part, I doubt that his advisor will allow him to do this before publication. Open science has not reached that level yet. You still have to publish your research somewhere to get noticed and get cited. I doubt Nature or any other journals will accept a paper with a blog article as one of the references.

-- Well, keep an eye on this. Certainly a brillant idea!

--Should I also blog my thesis when I write it? Maybe --

Friday, April 4, 2008

smoking leads to lung cancer??

Have you been told or been thinking that smoking causes lung cancer? Well, this is not exactly clear from recent studies.

An association between a genetic variation on chromosome 15 (15q24/15q25.1) and risk of lung cancer was found by three independent studies. But the link to smoking, or more scientifically nicotine dependence, is not clear. Some of the studies are still preliminary. So it is still possible that people with 'abnormal' chromosome 15 tend to be affected by environment - smoking - more easily and therefore are more likely to develop cancer in the long run.

We'll see if further studies with more - up to millions of - people can make it clear.

The papers mentioned here are:
  1. Hung, R. J. et al. Nature 452, 633–637 (2008).
  2. Thorgeirsson, T. E. et al. Nature 452, 638–641 (2008).
  3. Amos, C. I. et al. Nature Genet. doi:doi: 10.1038/ng.109 (2008).

Thursday, April 3, 2008

Dynamic "cold" genetic material

Two recent papers changed the concept of "cold" DNA, the so called heterochromatin.

1. Proliferation-dependent and cell cycle–regulated transcription of mouse pericentric heterochromatin
The Journal of Cell Biology, 2007; Vol. 179, No. 3, 411-421

Cell cycle control of centromeric repeat transcription and heterochromatin assembly

Genetic information resides in DNA in most organisms. DNA is packed into chromatin and stays in nucleus of a cell. There are two different package status: one loose and one tight. These are the original concepts of euchromatin and heterochromatin coined by German botanist Heitz 80 years ago. He proposed that heterochromatin reflects a functionally inactive state of the genome (all DNA information of an organism). Decades of research have generally been supportive to this idea. Heterochromatin is generally gene poor, highly packed, late replicating, and has a very low recombination rate. Modern molecular hallmarks of heterochromatin generally include heterochromatin protein 1 (HP1, swi6 in fission yeast) and methylation of histone H3 lysine 9 (H3K9) by histone methyltransferase (HMTase) (suv39, clr4). On the whole, Heitz’s original “inactive state” hypothesis still holds until recently.

The first paper demonstrated cell-cycle-specific transient disruption and transcription of mouse pericentric heterochromatin. It shows that mammalian pericentric heterochromatin is transcribed by RNA polymerase II twice during the cell cycle. A heterogeneous population of short RNAs (about 150 bp) is generated during mitosis, while a longer (mostly >1 kb) population is produced in late G1 and early S phase. Cell cycle regulation of pericentric transcription does not require Suv39h1,2-dependent chromatin modification, but it does require passage through "Start" in G1 phase. Future studies will determine whether these mammalian pericentric transcripts are important for heterochromatin formation as they are in fission yeast, as reported in the second paper.
in this paper, at regions serving as RNAi-dependent heterochromatin nucleation centers in fission yeast (in the pericentric, mating-type, and telomere regions), heterochromatin is abundant during G2 but greatly reduced during M, G1 and S phases. Heterochromatin reduction in M, G1 and S is correlated with phosphorylation of histone H3 on serine 10 (H3S10) and with binding of condensins. Genetic analyses show that condensin binding (in M and G1) and methylations of H3K36 (in S) and H3K9 (in G2) all contribute toward proper heterochromatin formation in G2 and toward proper regulation of transcription of the RNAi-dependent nucleation centers during S phase.

Because centromere /heterochromatin defect is almost the most common feature of cancers, this transcription could be the most basic factor during cancer formation. It might be that this transcription leads centromere structure problem, which leads to chromosome segregation defect which cause gene mutation etc. and cancer eventually.

Wednesday, April 2, 2008

Combining Genomic and Clinical Data for Cancer Therapy

This just came out today. So check it out.

An article entitled " Gene Expression Signatures, Clinicopathological Features, and Individualized Therapy in Breast Cancer" in the Journal of the American Medical Association, retrospectively studied traditional diagnostic standards of breast cancer outcomes — such as patient age, tumor size, and so on — and information about gene expression by modern genomic technology in a thousand breast cancer tumors. The conclusion is appealing: Gene expression patterns can, indeed, define subgroups of women with different prognoses and treatment responses.

“The combination of these two methods, one of which uses the clinical description of a patient’s breast cancer and the other which looks at gene expression at the molecular level in a patient’s tumor, may allow us to [match drugs with patients] with unprecedented accuracy,” senior author Anil Potti, an investigator at Duke University, said in a statement.

Traditionally, breast cancer evaluation is based on factors the so called TNM classification system, such as the patient’s age, tumor size, the level of lymph node involvement, and the degree of metastasis. These clinicopathological features could be employed to make predictions about clinical outcomes and help doctor’s to determine whether adjuvant cancer therapies such as chemotherapy or radiation therapy are warranted or necessary for different patients. However the estimation simply based on these factors are not always meaningful, e.g. it tends to overestimate cancer recurrence in younger patients.

To determine whether genomic data can provide additional information, the researchers studied women with early-stage breast cancer who had been followed for on average 11 years after initial assessment.
Indeed, the researchers did find that "molecular traits of patients in the poor prognostic clusters were highly specific and distinct from those of the good prognostic carriers”.

As the authors pointed out, identifying these subgroups may not only refine predictions about patient outcomes, it also provides information about patients’ underlying biology and the tumor microenvironment. That’s because gene expression patterns reveal different genetic pathways that are activated or silenced in different tumors during the long tumor formation progress.
For instance, low expression of cancer risk genes, chromosomal instability, and so on predict good outcome. However, high expression of genes associated with oncogenic pathway activation and wound healing etc. tend to be associated with poor outcome. Some genetic signatures also might indicate different responses to chemotherapy.

As wrote in an accompanying editorial in the same issue of JAMA, by Northwestern University researchers Chiang-Ching Huang and Markus Bredel, “This is one of the largest studies in human cancer showing the ability of gene expression profiles to improve risk stratification beyond established risk assessment algorithms that take into account clinicopathological variables”. This study “demonstrates the potential value of using microarray-based gene signatures to refine outcome predictions.”