User talk:Fubar: Difference between revisions

From genomewiki
Jump to navigationJump to search
No edit summary
 
(27 intermediate revisions by one other user not shown)
Line 1: Line 1:
'''CentOS notes'''
'''Things worth remembering that might be useful for other UCSC mirror builders'''


I just finished a limited install on a x86_64 CentOS 4.3 machine. There are some gotcha's worth knowing about.
== Notes ==
I only grabbed hg18.


[[Browser_Installation]] was very helpful.
Galaxy notes: [[Fubar:GalaxyNotes]]


Firstly, downloading the necessary parts of gbdb takes a very, very long time. There's a lot there.
Some random notes on installing a mirror under CentOS: [[CentOS_notes]]
Secondly, I gave up after a day of trying to get everything working anywhere but at the root of the website. All sorts of hardcoded references to ../trash and such like. Once I gave up and reverted to /var/www/html as the root, everything seemed to work better...
Thirdly, rsyncing the mysql directories directly works a treat. Just remember that the files become unavailable while they're being updated, but re-starting the rsync will work fine.


The following should work - they worked for me.
Some notes on adding a custom track: [[Custom_track]]
<tt>
export  MYSQLINC='/usr/include/mysql'
export MYSQLLIBS='/usr/lib64/mysql/libmysqlclient.a -lz -lcrypto -lssl -lm -lnsl'
export GLOBAL_CONFIG_FILE=/var/www/cgi-bin/hg.conf
export HGCGI=/var/www/cgi-bin
</tt>


If you have SE Linux enforcing on, you need to do something like:
== Code of potential use for mirror sites ==
<code>chcon -R  -u system_u -r object_r -t httpd_sys_content_t /var/www</code>
so apache can run cgi's and generally do useful things.


Strongly recommend that you do a cvs checkout of the source - that fixed a couple of misfeatures I found using the jksource.zip file.
The ucsc wiki can't set a cookie for your local mirror, so here's an example of
LDAP authentication for local mirrors: [[Fubar:LDAP_auth]]
 
== Gripes ==
 
Wiggle tracks need a bar graph option - points or joined points can be very misleading!
Ah, bedGraph mentioned in ..kent/src/hg/makeDb/trackDb/README does the trick. It's an ordinary 9 column bed file with extra options including bars which are much more appropriate for (eg) -log10(p) values from an analysis eg. I am making custom tracks using (eg)
<code><pre>
track bdrabpctmincov
priority 2.100
shortLabel bdrabpctmincov
longLabel CAMP 2007 pbat additive model -log10(p) bdrabpctmincov 
visibility hide
type bedGraph
autoScale off
viewLimits 0:7
minLimit 0
maxLimit 7
maxHeightPixels 60:40:40
color 56,88,9
group phenDis
</pre></code>
 
== Solutions ==
I think you may be confusing "wiggle tracks" here with the "chromGraph" tracks in the
"Genome Graphs" function of the browser.  Wiggle tracks are by default bar graphs.
The width of the bar in one of these graphs is the "span" of the data.  Each specified
data point applies to "span" number of bases.  See also, the discussion about
the proper usage of "span" in [[Wiggle_BED_to_variableStep_format_conversion]].
 
On the other hand, the "Genome Graphs" "chromGraph" data is completely different
than wiggle tracks.  In the chromGraph format, each specified data point is merely
a point and has no defined span to apply to a number of bases.  In this graph,
each point is simply connected to the next point with a line.  It is a line
graph, not a bar graph.  The chromGraph data does not have the ability to
define how large to make bars for a bar graph.  Certainly something to think
about for future improvements, but the purpose of chromGraph data is different than wiggle tracks.
 
Also, your note about mentions the bedGraph wiggle format. Please note the discussion
of the drawbacks of the bedGraph wiggle format in the above mentioned
link: [[Wiggle_BED_to_variableStep_format_conversion]]
 
[[User:Hiram|Hiram]] 09:12, 2 July 2007 (PDT)

Latest revision as of 16:12, 2 July 2007

Things worth remembering that might be useful for other UCSC mirror builders

Notes

Galaxy notes: Fubar:GalaxyNotes

Some random notes on installing a mirror under CentOS: CentOS_notes

Some notes on adding a custom track: Custom_track

Code of potential use for mirror sites

The ucsc wiki can't set a cookie for your local mirror, so here's an example of LDAP authentication for local mirrors: Fubar:LDAP_auth

Gripes

Wiggle tracks need a bar graph option - points or joined points can be very misleading! Ah, bedGraph mentioned in ..kent/src/hg/makeDb/trackDb/README does the trick. It's an ordinary 9 column bed file with extra options including bars which are much more appropriate for (eg) -log10(p) values from an analysis eg. I am making custom tracks using (eg)

 
track bdrabpctmincov
priority 2.100
shortLabel bdrabpctmincov
longLabel CAMP 2007 pbat additive model -log10(p) bdrabpctmincov  
visibility hide
type bedGraph
autoScale off
viewLimits 0:7
minLimit 0
maxLimit 7
maxHeightPixels 60:40:40
color 56,88,9
group phenDis

Solutions

I think you may be confusing "wiggle tracks" here with the "chromGraph" tracks in the "Genome Graphs" function of the browser. Wiggle tracks are by default bar graphs. The width of the bar in one of these graphs is the "span" of the data. Each specified data point applies to "span" number of bases. See also, the discussion about the proper usage of "span" in Wiggle_BED_to_variableStep_format_conversion.

On the other hand, the "Genome Graphs" "chromGraph" data is completely different than wiggle tracks. In the chromGraph format, each specified data point is merely a point and has no defined span to apply to a number of bases. In this graph, each point is simply connected to the next point with a line. It is a line graph, not a bar graph. The chromGraph data does not have the ability to define how large to make bars for a bar graph. Certainly something to think about for future improvements, but the purpose of chromGraph data is different than wiggle tracks.

Also, your note about mentions the bedGraph wiggle format. Please note the discussion of the drawbacks of the bedGraph wiggle format in the above mentioned link: Wiggle_BED_to_variableStep_format_conversion

Hiram 09:12, 2 July 2007 (PDT)