Introduction
　

Web Interface

ASAP II can be searched by several different criteria such as gene symbol, gene name and ID [UniGene, GenBank etc.]. The web interface provides 7 different kinds of views: (I) user query, UniGene annotation, orthologous genes and genome browsers; (II) genome alignment; (III) exons & orthologous exons; (IV) introns & orthologous introns; (V) alternative splicing; (IV) isoform and protein sequences; (VII) tissue & cancer vs. normal specificity. ASAP II shows genome alignments of isoforms, exons, and introns in UCSC-like genome browser (See A in following figure). Users can easily navigate among all the views by clicking links of interest. Alternative and constitutive exons are highlighted in red and blue, respectively. All alternative splicing relationships with supporting evidence information (See B in following figure), types of alternative splicing patterns (See C&D in following figure), and inclusion rate for skipped exons (See E in following figure) are listed in separate tables. Users can also search human data for tissue- and cancer-specific splice forms at the bottom of the gene summary page (See F in following figure). We report p-values for tissue-specificity as log-odds (LOD) scores, and highlight the results for LOD >= 3 and at least 3 EST sequences. See User's Guide for more comprehensive information.

Major output page for ASAP II. (A) Part of ASAP II genome browser showing isoform, exon and intron alignments. Alternative and constitutive exons & introns are denoted in red and blue, respectively. (B) List of all alternative splicing relationships. (C) Alternative donor and acceptor events. (D) Exon skipping events. (E) Exon inclusion rate for skipped exon. (F) Human EST library classification, tissue and normal vs. cancer specificity. High confident tissue-specific and cancer-specific alternative splicing relationships (LOD >= 3, at least 3 EST evidences) will be highlighted in red.

Statistics for ASAP II

^* Genome Assembly sequences were downloaded from UCSC genome browser except for Yellow fever mosquito, which was downloaded from Enesmbl genome browser
^** Alternative spliced genes (%) = No. of alternatively spliced clusters / No. of spliced clusters

53 % of human and mouse multi-exon genes are detected to contain alternative splicing. Focusing on genes with at least mRNA, 75 % and 60 % of human and mouse multi-exons genes were detected to contain alternative splicing (see ASAP II website for details). Due to limited mRNA and EST coverage (Fugu and honeybee) and incomplete genome assembly (Fugu, Ciona, and yellow fever mosquito), number of mapped clusters (Ciona, 9 %; 1373 out of 15587) or alternatively spliced clusers (24 for Fugu, 98 for Ciona, 57 for honeybee, 87 for and yellow fever mosquito) can be significantly lower than expected, these data cannot be considered comprehensive. 19 ~ 26 % of fruit fly, western clawed frog, chicken, rat, and cow multi-exon genes were detected to contain alternative splicing.

Click here to download full statistics for ASAP II

UniGene release date and Statistics for Alternative Splicing Events
　

Organism	UniGene Build	Release Date	Exon Skipping	Alternative Donor	Alternative Acceptor	Mutually Exclusive	Total Alternatively Spliced Cluster
Human	#188	04 Dec 2005	7868	3307	2408	1365	11717
Mouse	#151	02 Oct 2005	4147	1870	1277	321	8711
Rat	#148	04 Oct 2005	1238	448	293	55	3378
Western clawed frog	#28	13 Oct 2005	878	302	229	31	2349
Chicken	#29	12 Dec 2005	753	254	199	30	2154
Cow	#74	14 Dec 2005	167	483	279	45	3008
Dog	#13	19 Aug 2005	328	75	38	16	951
C. elegans	#25	25 Sep 2005	192	95	90	10	763
Ciona	#18	02 Nov 2005	24	15	14	1	98
Zebrafish	#89	05 Dec 2005	448	223	138	8	1577
Fruit fly	#40	24 Nov 2005	409	263	109	35	1841
Fugu	#3	30 Dec 2004	4	1	2	0	24
Yellow fever mosquito	#1	04 Jan 2006	12	11	7	0	87
Honeybee	#5	31 Jul 2004	11	10	6	1	57
African malaria mosquito	#31	17 Sep 2005	68	50	36	11	605

All UniGene database was download in January 2006. Due to UniGene updating frequency, release date would be older than January 2006. Current release of ASAP II is version JAN06. We will update ASAP II database if total number of available sequences is significantly increased. Right side of table shows statistics for alternative splicing events. Exon skipping events are most common events.

Above figure shows segment of multigenome alignment. As you can see, human exon is well conserved for mouse, rat, rabbit, dog, elephant and opossum. And, splice site consensus is all canonical, AC/CT (GT...AG if reverse complement). By comparing all splice sites within ASAP II database, we constructed ortholous exons and introns database.

Statistics for Orthologous Exons and Introns

^* Only orthologous exons and introns that have two exact matches of both canonical splice sites (U1/U2 and U11/U12).
^* List of species used in multigenome alignments is available at UCSC genome browser.

66 % (85673 out of 129981) for human and 77 % (81296 out of 105260) for mouse internal exons have at least one orthologous exons. 52 % (100447 out of 193024) for human and 69 % (97371 out of 141285) for mouse canonical introns have at least one orthologous introns. Most of orthologous exons and introns were from human and mouse orthologs due to larger number of mRNA and EST sequences than other genomes. 56 %, 37 %, and 47 % of chicken (galGal2), zebrafish (danRer3), and western clawed frog (xenTro1) internal exons have at least one orthologous exons and 49 %, 33 %, and 41 % for orthologous introns. Because a set of genome assemblies used for multigenome alignments is different from ASAP II calculation for chicken, zebrafish and western clawed frog (see Table 1 for details), numbers of orthologous exons and intron cans be decreased.

Multiple alignment of splice sites is extraced using Pygr and its phylogenetic tree is generated by UCSC Phylogenetic Tree GifMaker on the fly.

We found 1709 high-confidence (LOD >= 3) tissue-specific alternative splicing relationships from 960 genes (Click Here to download all high confident tissue-specific relationships), and 273 high-confidence (LOD >= 3) cancer-specific relationships from 198 genes (Click Here to download all high confident cancer-specific relationships). The largest categories of tissue-specific splice forms were identified from brain/nerve, testis, skin, muscle, and lymph (Click Here to download statistics for tissue-specific genes).  Users can download all EST library classification and LOD calculation results from ASAP II download page and mine their own experimental candidates.

After one uploads MySQL tables in download page, all high confident alternative relationships can be retrieved by following SQL syntax.

mysql> select * from LOD_Tissue_hg17 where LOD >= 3 and n_s1_tissue >= 3 order by cluster_id;
mysql> select * from LOD_Cancer_hg17 where LOD >= 3 and n_s1_tissue >= 3 and tissueId = 1 order by cluster_id;
mysql> select count(t1.tissueId), t1.tissueId, t2.tissueTerm from LOD_Tissue_hg17 as t1, EST_Library_Classification_Tissue_Term_hg17 as t2 where t1.LOD >= 3 and t1.n_s1_tissue >= 3 and t1.tissueId = t2.tissueId group by t1.tissueId;

1. How to upload ASAP II into your own database

After creating your own database, e.g. SPLICE_JAN06, you can upload all ASAP II database into your database server.

mysql> create database SPLICE_JAN06;

Download MySQL file in download page. And, uncompress using "gzip -d" command. You can upload using "mysql" command.

$ mysql SPLICE_JAN06 < alt_splice_hg17.sql

You can check uploaded database using SQL syntax.

mysql> desc alt_splice_hg17;
+--------------+---------------------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+--------------+---------------------------+------+-----+---------+-------+
| splice_id_1 | int(11) | NO | PRI | 0 | |
| n_est_obs_1 | int(11) | YES | | NULL | |
| n_mrna_obs_1 | int(11) | YES | | NULL | |
| gen_start_1 | int(11) | YES | | NULL | |
| gen_end_1 | int(11) | YES | | NULL | |
| splice_id_2 | int(11) | NO | PRI | 0 | |
| n_est_obs_2 | int(11) | YES | | NULL | |
| n_mrna_obs_2 | int(11) | YES | | NULL | |
| gen_start_2 | int(11) | YES | | NULL | |
| gen_end_2 | int(11) | YES | | NULL | |
| cluster_id | varchar(12) | NO | MUL | | |
| is_novel | enum('no','yes') | YES | | NULL | |
| evidence | enum('single','multiple') | YES | | NULL | |
+--------------+---------------------------+------+-----+---------+-------+
13 rows in set (0.00 sec)

In order to analyze ASAP II database using Pygr, you have to upload all tables in ALTSPLICE and ISOFORM at download page. Suffix of each table is genome assembly name, e.g. *_hg17.

2. How to install Pygr and Tutorials

Prerequisites: Python 2.2 or higher (Python 2.3 or higher recommended), Pyrex, MySQL client & Python MySQLdb module

You can download Pygr at http://sourceforge.net/projects/pygr/, Pygr can be installed by standard python installation method. (For distribution package, we already generate C files from Pyrex *.pyx. You don't need Pyrex unless you want to do change *.pyx files). Connection to MySQL database is essential, thus you need both MySQL client and MySQLdb python module to work with ASAP II database.

$ python setup.py install

Check whether Pygr is installed correctly.

$ python
Python 2.4.2 (#2, Sep 30 2005, 14:23:11)
[GCC 3.2.2 20030222 (Red Hat Linux 3.2.2-5)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import pygr

Pygr was presented in a software demo at ISMB2005 and ISMB2006. For Pygr 0.5, see this presentation for details.

3. Graph Query Examples

After you uncompress Pygr, you can see bunch of test modules in tests directory. Test module is based on ASAP or ASAP II database (with *_jan06.py). For Pygr graph query test, test_jan06.py is successfully tested using ASAP II database. You may need to change database name and table name in lldb_jan06.py. For human, it would take about 40 minutes (and about 1GB memory) to upload all alternative splicing relationships in ASAP II database. Results of test_jan06.py is test_jan06.log. Compare log file with your own results.

4. Multigenome Alignment Query using Pygr NLMSA

First thing to do is to download multigenome  alignment at UCSC genome browser. Many multigenome alignments are available, but you have to download same version with ASAP II database (see Statistics table above).

After you finish downloading and uncompress in your directory, you can generate Pygr NLMSA using createdb.py script. You may have to change chromosome names in the script.

In Pygr tests directory, you can find alt35_jan06.py. This test script is to extract splice sites from multigenome alignments. The output of that script is as follows.

1 Hs.99886 ss1 panTro1 100.0 100.0 GT GT
1 Hs.99886 ss1 monDom2 50.0 100.0 GT GA
1 Hs.99886 ss1 loxAfr1 100.0 100.0 GT GT
1 Hs.99886 ss1 dasNov1 100.0 100.0 GT GT
1 Hs.99886 ss1 rheMac2 100.0 100.0 GT GT
1 Hs.99886 ss1 echTel1 100.0 100.0 GT GT
1 Hs.99886 ss1 bosTau2 100.0 100.0 GT GT
1 Hs.99886 ss1 canFam2 100.0 100.0 GT GT
1 Hs.99886 ss1 rn3 50.0 100.0 GT AT
1 Hs.99886 ss1 mm7 50.0 100.0 GT CT
1 Hs.99886 ss2 panTro1 100.0 100.0 GT GT
1 Hs.99886 ss2 echTel1 100.0 100.0 GT GT
1 Hs.99886 ss2 loxAfr1 100.0 100.0 GT GT
1 Hs.99886 ss2 dasNov1 100.0 100.0 GT GT
1 Hs.99886 ss2 oryCun1 100.0 100.0 GT GT
1 Hs.99886 ss2 rheMac2 100.0 100.0 GT GT
1 Hs.99886 ss2 monDom2 100.0 100.0 GT GT
1 Hs.99886 ss2 bosTau2 100.0 100.0 GT GT
1 Hs.99886 ss2 canFam2 100.0 100.0 GT GT
1 Hs.99886 ss2 rn3 100.0 100.0 GT GT
1 Hs.99886 ss2 mm7 100.0 100.0 GT GT
1 Hs.99886 ss3 panTro1 100.0 100.0 AG AG

The python script, alt35_jan06.py is only for ASAP II database. If you want general Pygr NLMSA feature, multigenome alignment query, see ISMB2006 presentation PDF for details.