Blip: Modules

Modules

The Blip modules have been documented using PrologDocNG, by Salvador Fandino (thanks Salvador!). You can access the autogenerated HTML (one file per module) via this page. Each HTML file has a link to the source at the bottom.

Introduction

Blip consists of different prolog modules covering different areas of knowledge withins the life sciences - sequences, analysis, ontologies, phylogenetics, pathways, etc. These modules are largely independent - you may only need to use one of the modules below for your research.

The modules are in two sections: MAIN and EXT. The MAIN modules are all in pure prolog and should run without problems. The EXT modules use the SWI-Prolog foreign language interface to use other libraries (currently C, but java and C++ may be on the cards for the future).

Of the Main modules, some are bio modules (containing code specific to some subdomain of biology), and some are generic modules (which may be of interest to other SWI-Prolog users regardless of their interest in bioinformatics). The latter may eventually be split off from blip and go into independent packages.

It may also be an idea to become familiar with some of the standard SWI-Prolog packages

Main Bio Modules

These modules contain specific biological semantics. There are three kinds of modules:

data or domain model modules, which define a set of data predicates (facts) and view predicates over these facts, which can be imported from some external file (see the io module). Data modules should be named with the suffix _db. Data modules can also be thought of as relational database schemas (albeit with added behaviour).
utility modules, which do not depend on any data module, and do not require any data to be stored in the database.
bridge modules, which map one data model to another. Blip does not insist on everything conforming to one uber-model. For example, sometimes it is convenient to model genomic features using the seqfeature_db model, and other times it is more convenient to model them as instances of SequenceOntology classes using the ontol_db module or SemWeb

Module:seqfeature_db [src]

usable

A model of biological sequence features. Maps well to standard bio models, in particular GFF3, Chado and Chaos-XML
This module also uses bioseq and range

This data module can use the following inter-model bridges

Module:ontol_db [src]

stable

Models classes and instances. This model maps well to both OBO and OWL models

This data module can use the following inter-model bridges

Module:sb_db [src]

usable, incomplete

Models pathways and interactions. Currently the model is very SBML-y (level 1 and 2) but it may change. Little support for quantative stuff as yet.

This data module can use the following inter-model bridges

Module:pheno_db [src]

in use

Genotype and phenotype data predicates

This data module can use the following inter-model bridges

Module:phylo_db [src]

test

Phylogenetic tree data

This data module can use the following inter-model bridges

parser_nhx [src]

Module:seqanalysis_db [src]

new

Models results of sequence similarity searches, eg blast results

Module:taxon_db [src]

usable

Models a taxonomic tree of organisms. Fairly simple and stable. Some overlaps with phylo_db

This data module can use the following inter-model bridges

taxon_bridge_to_ontol [src]

Module:io [src]

usable

handles input/output - mostly loading from external files into the in-memory prolog database

Module:biotable [src]

usable

Metadata about many common tabular bioinformatics file formats, for use in conjunction with the table module

Module:bioprolog_util [src]

usable, but not stable

hodgepodge of useful predicates. Some predicates may migrate to more specialized modules in future

Module:bioseq [src]

usable, but currently very limited functionality

Provides some handy predicates for handling nucleic acid and amino acid sequences. Does not assume any model

This module makes use of:

Module:gencode [src]

usable

genetic codes (translation tables) for different species classes

Module:iupac [src]

usable

Module:range [src]

usable, interface subject to change

Operations on linear ranges. Does not assume any particular model (predicate arguments are typically compound terms representing range "objects")

Parsers

File format parsers - these convert from flat files conforming to some grammar into a list of prolog terms. These prolog terms may then later be converted to datalog facts in one of the db modules above; or the prolog terms may be an end in themselves

Most parsers will be prolog DCGs. This means they may be reversible - from prolog terms to flat file syntax

Module:parser_nhx [src]

stable

Parses New Hampshire and New Hampshire Extended format (phylogenetic trees). See also phylo_db

Module:parser_obo [src]

beta

Parses obo text format. Not currently ysed. For now, the io module will use go-perl behind the scenes to convert this. See also ontol_db

Module:parser_nlp [src]

stable

Parses the output of most NLP lexical parsers; eg Stanford lexparser

Generic Modules

General purpose modules, not specifically tied to ontologies, bioinformatics or biomedical applications

(These modules may also be of interest to non-bio programmers)

Module:plunit [src]

usable

Unit testing

Module:graph [src]

incomplete, usable

Generic graph operations - closure, least common ancestors etc. Higher order - can be used with different db modules.

Module:rdb_util [src]

usable

SQL database interface, layered on top of the ODBC interface

Module:google [src]

usable

Wrapper for the Google SOAP API - enables google queries to be performed from prolog

Module:tabling [src]

usable, limited

Simple memoization of predicates

Module:xml_writer [src]

usable

Maps prolog term syntax XML to actual XML

Module:xml_transform [src]

incomplete, usable

XPath-like query language and declarative transforms between XML formats and flat predicates

Module:serval [src]

usable, in flux

A lightweight logical-functional framework for rapidly deploying web applications

Module:graphviz [src]

usable

Simple predicates for generating graphviz dot files from compound terms

Bio Ext Modules

These modules live in the "ext/" directory, and are not installed by default. To install, descend into this directory and follow the instructions there.

The idea is that this directory will contain modules that use the SWI foreign language interface to call C libraries for tasks that may be slow in Prolog. So far this just has one proof of concept module

Module:squid [src]

not complete, but still usable

Uses Sean Eddy's SQUID C library for parsing sequence files (various formats), and providing an alternate implementation of revcomp/2