Thursday, May 23, 2013

PHP Caused Segmentation Fault - php-pecl-ssh2

I have being trying to debug a php code for the last couple of days. Coming from Python, it was frustrating not to be able to have nice debugging tools available. I ended up installing latest ssh2 php module to fix this. It seems that PHP :: Bug #63192 was causing this. I tried Eclipse PDT, Xdebug and var_dump, but couldn't figure out how to fix this. 

Here are some technical details, in case you are facing the same problem:
$ rpm -qa|grep ssh2
libssh2-devel-1.2.2-11.el6_3.x86_64
libssh2-1.2.2-11.el6_3.x86_64
php-pecl-ssh2-0.11.0-7.el6.x86_64
  
$ more /etc/issue
CentOS release 6.3 (Final)
Kernel \r on an \m

wget http://pecl.php.net/get/ssh2-0.12.tgz
tar -xzvf ssh2-0.12.tgz 
cd ssh2-0.12
yum install automake make php-devel libtool openssl-devel gcc++ gcc
phpize
yum install gcc php-devel php-pear libssh2 libssh2-devel
./configure 
make
make test
cp modules/ssh2.so   /usr/lib64/php/modules/
service httpd restart

Friday, October 29, 2010

Medicine Is Personal Again

I watched excellent NIH videocasts on personalized medicine called Comparative Effectiveness and Personalized Medicine: An Essential Interface. Here are the links to recordings: day 1 and day 2 that I highly recommend. The conference was very informative and well done. It started with introduction on how personal genome sequencing can influence the future of medicine. One of the interesting questions asked was about using the term personalized medicine. Doctors where doing personalized medicine for ages; a routine blood pressure measurements gives a personalized info, so it seems that genetic testing won't be that different.

It was interesting to hear the story about Gefitinib (marketed as Iressa). PharmGKB has annotations that shows specific mutations associated with improved progression free survival in gefitinib-treated non-small-cell lung cancer patients. There were other examples presented to show how personalized genome helped directing clinical trials towards more useful results. The search for genetic variant in http://clinicaltrials.gov returned 421 studies as of this writing.  There were also questions raised about how affective existing drugs would be if one can guarantee adherence to drug regimen.

I also liked the presentation about http://www.armyofwomen.org. It seems that one of the major problems in drug discovery today are clinical trials that are too expensive and hard to do.

Overall, I learned a lot from the conference and I'm very grateful to NIH for hosting the videocasts.

BioChem Related Feeds

Feeds from Journals