PHILM2Web

PHILM2Web provides a high-throughput, literature-mined database of macromolecular host–pathogen interactions to support systematic analysis of pathogen entry, immune evasion, spread, and multiplication within hosts.

Key Features:

• High-Throughput Literature Mining: The PHILM (Pathogen-Host Interactions by Literature Mining) algorithm automatically extracts HPI data from PubMed abstracts and processed over 608,972 PubMed entries that mention both host and pathogen entities.
• Comprehensive Database: The database contains 23,581 generic HPIs involving 157 host and 403 pathogen organisms, extracted from 11,609 abstracts.
• COVID-19 Specific Data: An additional 25,796 PubMed abstracts were processed using the same query parameters as the COVID-19 Open Research Dataset, yielding 257 specific HPIs between 19 host and 31 pathogen organisms from 167 abstracts.
• Integration with Existing Knowledge: Interactions are compiled from generic databases and curated experimental studies to augment literature-mined HPIs.
• Extraction Precision: The methodology emphasizes automated extraction using PHILM to achieve high precision in identifying relevant host–pathogen interactions.

Scientific Applications:

• Pathogen Entry and Immune Evasion Studies: Enables exploration of molecular interactions underlying how pathogens breach host defenses and evade immune responses.
• Drug Discovery and Development: Supports identification of potential therapeutic targets by mapping interactions critical to pathogen survival and replication.
• Vaccine Research: Informs antigen selection and mechanistic understanding of immune responses by providing detailed HPI data relevant to vaccine design and evaluation.

Methodology:

Automated extraction of host–pathogen interactions from PubMed abstracts using the PHILM algorithm, processing 608,972 PubMed entries to obtain 23,581 HPIs from 11,609 abstracts, and processing an additional 25,796 PubMed abstracts with the same query parameters as the COVID-19 Open Research Dataset to obtain 257 COVID-19–related HPIs; extracted interactions are integrated with data from generic databases and curated experimental studies.