ChEMBL 08 (other versions as well) dumped the data as mysql or oracle format, however, all chem2bio2rdf bases on postgres, so we need to upload chembl into postgres and then integrate with other chemogenomics datasets.

I tested several scripts (mysql2pgsql.perl, ) which claim to convert mysql dumped file to postgres readable format, but I failed. mysql2pgsql.perl is crashed due to segmentation fault, the data converted from the latter script came out a lot of errors. So I decided to manually import it. Here are the steps.
1) import mysql dumped file into mysql.
2) import the tables in mysql into access as linked table using odbc, (you have to create DSN linking to mysql and another DSN linking to postgres
3) export the table to postgres using odbc.
4) one table 'activities' is too big to manage by odbc in access, so I export it to a text file and use copy function in postgres to upload
5) compound table is too big due to the molfiles, since we don't need this column, simply deleted it. '
6) in the compound table, convert the compound into CID using our pipeline protocal.
7) upload compound, ID: 1793, 263053 failed, cause the data has bad format.