Introduction
PMC-AB aims to mimic CiteAB's functionaliy as an Antibody citation search engine.
Currently we search articles on EuropePMC.
This software is packaged as a Python module.
Workflow
For a given input that uniquely identifies a commercially available antibody, this program generates EuropePMC-compatible search queries that will return articles that cite that antibody.
Queries are generated by replacing previously-known templates that represent ways that a antibody can be referenced in academic texts
Input
Antibody data: SKU, CLONE_ID, MANUFACTURER, ``
Query templates
$SKU $MANUFACTURER
#$TARGET*$CLONE
#$MANUFACTURER AND $CLONE
#$SKU AND $MANUFACTURER
"$TARGET \(Clone #$CLONE, $MANUFACTURER"
"$TARGET antibody \($CLONE; $MANUFACTURER"
"$MANUFACTURER clone $CLONE"
"$MANUFACTURER, clone $CLONE"
Built queries
Applying antibody information to our query templates yield a list of independent queries:
"CD4 \(Clone #RM4-5, BD Biosciences"
"CD4 \(Clone #RM4-5, BDBiosciences"
"CD4 \(Clone #RM4-5, BD-Biosciences"
"CD4 antibody \(RM4-5; BD Biosciences"
"CD4 antibody \(RM4-5; BDBiosciences"
"CD4 antibody \(RM4-5; BD-Biosciences"
"BD Biosciences clone RM4-5"
"BDBiosciences clone RM4-5"
"BD-Biosciences clone RM4-5"
"BD Biosciences, clone RM4-5"
"BDBiosciences, clone RM4-5"
"BD-Biosciences, clone RM4-5"
We can combine independent queries with OR directives so we can query all of them in a single API call:
(550280 BD Biosciences) OR (550280 BDBiosciences) OR (550280 BD-Biosciences) OR ("CD4 \(Clone #RM4-5, BD Biosciences") OR ("CD4 \(Clone #RM4-5, BDBiosciences") OR ("CD4 \(Clone #RM4-5, BD-Biosciences") OR ("CD4 antibody \(RM4-5; BD Biosciences") OR ("CD4 antibody \(RM4-5; BDBiosciences") OR ("CD4 antibody \(RM4-5; BD-Biosciences") OR ("BD Biosciences clone RM4-5") OR ("BDBiosciences clone RM4-5") OR ("BD-Biosciences clone RM4-5") OR ("BD Biosciences, clone RM4-5") OR ("BDBiosciences, clone RM4-5") OR ("BD-Biosciences, clone RM4-5")
Validation
To validate search results, we compare to the .xlsx reference, which is based on CiteAB results.
Application execution two type of files:
- A
.csvrecord containing overall statistics:
| Antibody ID | Fulfillment Rate | Fulfillment rate PMC | False Positive Rate | Search Hit Count | Search N | Agreement N | Benchmark N (CiteAB) |
|---|---|---|---|---|---|---|---|
| BD Biosciences:550280 | 63.76% | 78.54% | 95.13% | 6952 | 6900 | 336 | 527 |
| Invitrogen:11-0041-82 | 35.95% | 39.77% | 96.31% | 4471 | 4472 | 165 | 459 |
| Abcam:ab183685 | 80.71% | 97.66% | 16.31% | 325 | 325 | 272 | 337 |
| BioLegend:100401 | 92.47% | 87.61% | 94.38% | 5467 | 5467 | 307 | 332 |
| Abcam:ab133616 | 77.62% | 93.33% | 29.75% | 316 | 316 | 222 | 286 |
| R&D Systems:AF1828 | 66.67% | 90.0% | 4.76% | 42 | 42 | 40 | 60 |
| R&D Systems:BAF1828 | 71.43% | 100.0% | 0.0% | 20 | 20 | 20 | 28 |
- A list of articles that were retrieved by the application but were not in the
.xlsxrecords, one per antibody.
Usage
Setup
Usual Python package setup:
pip install -r requirements.txt
pip install .Execution
To evaluate all antibody records and produce result records. Files will be created in the current directory, which can be any.
pmcab -a