This presentation demonstrates how to use Elsevier Text Mining to analyze a patient's microbiome profile from Aperiomics and interpret the results. Multiple Search is used to identify bacterial species from the patient's throat and stool that have been linked to her lung and bowel problems in medical literature. Species linked to her conditions include Streptococcus pneumoniae, Streptococcus pyogenes, Haemophilus parainfluenzae, and Fusobacterium nucleatum in her throat and Bacteroides vulgatus in her stool. Multiple Search also identifies protective species in her stool like Bifidobacterium longum and Faecalibacterium prausnitzii. Antibiotics and probiotics are then suggested
Goa Call Girl Service 📞9xx000xx09📞Just Call Divya📲 Call Girl In Goa No💰Advanc...
Patient microbiome analysis using Elsevier text mining
1. October 2019
Elsevier Life Science Solutions
Analysis of patient microbiome
using Elsevier Text Mining
Demonstrating value via real-world use cases
2. Thanks to Human Microbiome Project more and more literature data become
available linking different bacterial species from human microflora to various
human disease and conditions. Companies like Aperiomics, Eagle Genomics,
uBiome offer NGS-based services to individual patients to determine their
microbiome composition. The interpretation of their data to produce actionable
clinical outcomes still remains a challenge. This presentation shows how to use
Elsevier Text Mining to interpret microbiome profile of a patient, find plausible
pathogenic bacteria and treatments to optimize malfunctioning microbiome
3. Elsevier Text Mining is an easy to use text analytics solution that integrates
several disparate data sources to help address customers’ problems
Researchers’
Complex Questions
Novel Actionable Insights
& Comprehensive Answers
Supporting Additional Downstream Analyses
Technology and AnalyticsContent
• RELX provided
• Public Content
• Customer Content
• domain expertise
• state-of-the-art NLP
• flexible reporting and API access
Elsevier Text Mining Product Platform
APIUI
NLP toolkit
Search
Analytics
Visualization
Data Connectors
Content Ontology
Uploads / Content Integration
4. Overview of human microbiome (Wikipedia)
About 100 trillion bacterial cells from more than 1000 different bacterial species live in human body
5. Patient symptoms and diagnosis
27-year old woman with history of depression,
chronic bronchitis and occasional abdominal pain,
suggesting episodic inflammatory bowel disease
Microbiome profile for the patient was measured by Aperiomics Inc.
9. Which bacterial species from patient
throat have been linked to patient’s
lung problems in medical literature?
10. Multiple Search in Elsevier Text Mining solution
is an ideal tool to help find the answers
List of bacterial species
found in patient by
Aperiomics
List of patient’s
symptoms
Query looks for articles containing
semantic co-occurrence relations
between $X and $Y terms
Using Multiple Search one can
create a semantic co-occurrence
relationship matrix based on the
existing scientific literature to
quickly visualize publications
pertaining to the microbiome
results and subsequently dig
deeper into the specifics
11. Which bacterial species from patient stool
have been linked to patient’s enteritis-like
problems in medical literature?
12. Again, one can use Multiple Search
to create a semantic co-occurrence
relationship matrix based on the
existing scientific literature to
quickly visualize publications
pertaining to the microbiome
results and subsequently dig
deeper into the specifics
14. Finding antibiotics optimizing patient microbiome
Export search results from Elsevier Text Mining for further review and to make unique list of antibiotics
15. Antibiotics must inhibit pathogenic bacteria
but do not affect protective bacteria
Antibiotics having desired efficacy profile
are shown by red arrows
16. Finding probiotics for the patient
Which bacteria is known to inhibit enteritis and
depression but not present in patient microbiome?
17. 1. Run a search with Elsevier Text Mining to identify
semantic co-occurrences where bacteria inhibit in the
context of enteritis
2. Export results from Elsevier Text Mining to make a
unique list of bacteria species
Finding evidence for bacteria that inhibit enteritis
18. Use Multiple Search to find bacterial species inhibiting enteritis
which also inhibit depression
Already present in
patient’s stool
Already present in
patient’s stool