The PKSDB has been organised such that it has two distinct arms:

• The first allows the user to mine the modular polyketide synthase clusters charcaterized through experiments. The sequence and function data from 19 experimentally characterized modular PKS clusters has been compiled in PKSDB.
• This facility enables easy extraction of various domain sequences as well as linker sequences present in each of these 19 clusters along with the chemical structures of their products. Catalytic activity of reductive domains, active site residues,and substrate specificity are also marked out for each of these 19 clusters in the PKSDB.
• Click here to get the references for the 19 PKS clusters compiled in the PKSDB.
• The second aspect is the search facility that allows users to enter any test protein sequence and analyse it for potential polyketide domains. 
• Here is how these two interfaces have been put together in the database: 

• As can be seen, the left frame lists the nineteen characterized PKS clusters which have been comprehensively compiled in the PKSDB. The search facility is also in this frame. 
• Upon selecting a particular PKS domain cluster or in response to a query submitted by the user, PKSDB shows a pictorial depiction of the domain organisation as a webpage consisting of clickable links to each domain region, each linker region, each orf in the cluster and the chemical structure in case of a characterized PKS cluster. 
• Here is an example:

Key to domain images on the domain organisation page:
Each domain has been represented using a coloured circle with the name of the domain inscribed in it. The inter domain linker regions have been depicted as filled black lines connecting the domains, while the N- and C- terminal linkers have been represented as shaded lines in violet. All domain s within the same module have been depicted using a single colour, while different modules have different colours. 

Filled circles	Domain successfully identified by PKSDB
Dotted circle	Domain missed by PKSDB
Upper case letters	Catalytically active domain
Lower case letters	Catalytically inactive domain
Dotted and lower case	Non-functional missed domain.

• This page also leads to a color image of the product of the polyketide cluster chosen which enables immediate comprehension of the correspondence between each product and the constitution of the domains in the respective PKS cluster.
• As is well known, the reductive domains in modular PKS are optional, and therefore, PKSDB allows the user to assess the functionality of each reductive domain by simply looking at the pictorial depiction of the domain. Catalytically inactive domains are inscribed in lower case letters while active ones are in uppercase letter.
• For every domain represented on the domain organisation page, there is an image as described above, which when clicked will lead to the respective orf, domain or linker page.
• This page allows the user to analyse each domain in detail and compare it with all other domains of its kind. 

• Here is an example:

• This page can be used for comparison of homology between any two characterized PKS domains or between a characterized and a potential PKS domain ( i.e domain identified by PKSDB in a query ) in terms of (i) evalue  (ii) Percentage  sequence identity  (iii) Percentage sequence similarity.
• In addition to sequence homology comparisons, substrate specificities and even active site motifs of all domains can be compared since the PKSDB lists the substrate specificities of all modules along with the active site residues of each AT domain. 
• The domain page shown above has three forms to perform these functions:
• Selection of the first form on this page pops up a window with the sequence of the respective region in fasta format. 
• Using the second form, the user can choose to see which other characterized domain is closest or farthest from the given domain in terms of sequence homology. Various attributes have been rendered to make this search simple and easy. It takes you to a webpage which lists in the required order, all other domains satisfying the variables filled by the user. Each of the domains in this list will be a clickable link to the alignment of the domain in question with the selected domain. 
• Upon clicking the third form, user can view the alignment of the domain in question with any other characterized domain of its kind picked from any module of any of the 19 clusters listed on the left panel. 

• The query interface of PKSDB allows the user to identify the possible polyketide domains in a modular PKS cluster with unknown polyketide product or in a PKS-like unannotated cluster.
• Appropriate prediction rules have been incorporated to predict the specifcity of various AT domains for different starter and extender units. This interface has been added to the PKSDB as a step further in the direction of eventual, in silico predicition of polyketide products. 
• Here is how the query page appears in the DBASE:

• The user can submit for testing, at the same time, upto ten different protein sequences using this form. This has been incorporated keeping in mind that polyketide synthase protein complexes tend to occur in clusters. 
• Clicking on this page will lead to a form with user specified number of text boxes to enter the query sequence/ set of sequences.
• The submission of the query will lead to a domain organisation page as mentioned above with details of all potential polyketide domains in the query sequence or cluster.
• Even though the present version of the search tool cannot predict which orf in a given cluster may be the first in the biosynthetic process, but it can assist the user in predicting what chemical moeity a given orf would contribute to the final product. 
• In addition, as already mentioned above, it can predict the starter or extender unit specificity for the various AT domains. 
• It also extracts and lists the active site residues of the AT domains after an alignment of the query domain with the E.coli MCAT whose crystal structure has been determined. 
• Navigation from the domain organisation page is self-explanatory and can be done in the same manner as described above.