IMPC Release Notes

  • Release: 2.0
  • Published: 06 November 2014
Statistical Package
  • PhenStat
  • Version: 1.2.0
Genome Assembly
  • Mus musculus
  • Version: GRCm38
  • Number of phenotyped genes: 518
  • Number of phenotyped mutant lines: 535
  • Number of phenotype calls: 2,182

Data access


New lines added, some taken away

In data release 2.0, we now have 535 mouse lines with phenotype data that has passed quality control. This positive step demonstrates the majority of IMPC centers are now able to generate, capture, export, passed quality control (QC), and distribute high-throughput data for hundreds of new mouse lines. Going forward, we anticipate bi-monthly data releases.

In review of our processes, we noticed some lines passed QC in the previous data release that should not have. We have reverted the statuses of these lines to a pre-QC state, meaning they our visible on the portal but not being distributed via the API. The majority if not all of these lines should be available in the next data release.

Pre-QC vs Post-QC

In striking a balance between making data available to the public as soon as possible versus ensuring data is of the highest quality, the data coordination center (DCC) provides both "Pre-QC" and "Post-QC" data.

Pre-QC data is defined as phenotype data that has been uploaded to the DCC and has passed automated checks but has not been signed off by a data wrangler. Reasons for not signing off include not enough animals being tested, not all mandatory procedures have had data uploaded, or a known issue has been reported to a phenotyping center and is waiting to be resolved. Lines with pre-QC data can be found by filtering on "phenotype started" under "Gene" -> "IMPC phenotype status" on the search page.

Post-QC data is defined as phenotype data for a mouse line that has been signed off by a data wrangler. This means a mouse line has had enough animals tested for the core procedures and has passed quality control. In this phase of the project, we have restricted the number of core procedures that data is needed for as certain data types are more difficult to upload than others. Lines with Post-QC data can be found by filtering on "phenotype complete" under "Gene" -> "IMPC phenotype status" on the search page.

In this latest release, we present both pre and post QC data on the portal. Both sets of data are analysed using the same statistical analysis package. Phenotype associations made from Pre-QC data have orange graph links while post-QC data have blue links. Only phenotype associations made from post-QC data are available in the API and pushed to third parties.

Statistical Analysis

The same statistical analysis package version used in our last data release is used for this release, Version: 1.2.0. We anticipate use PhenStat 2.0 in our next data release.

Data Reports

Lines and Specimens

Phenotyping Center Mutant Lines Baseline Mice Mutant Mice
MRC Harwell 82 2,318 1,886
HMGU 16 511 244
ICS 30 525 473
WTSI 246 1,631 3,713
JAX 23 1,114 583
UC Davis 57 618 1,243
TCP 66 312 1,288
BCM 16 292 354

Experimental Data and Quality Checks

Data Type QC passed QC failed issues
categorical 1,173,776 0 1,764 *
unidimensional 988,022 992 * 16,928 *
time series 1,798,263 4 * 39 *
text 15,355 1,721 * 100 *
image record 10,765 0 0

* Excluded from statistical analysis.


Allele Types

Mutation Name Mutant Lines
Targeted Mutation 2 3
Targeted Mutation 1 77
Targeted Mutation e 16
Targeted Mutation b 231
Targeted Mutation a 211

Mouse knockout programs: EUCOMM,KOMP

Distribution of Phenotype Annotations



By Center

More charts and status information are available from iMits.

Phenotype Associations Overview

We provide a 'phenome' overview of statistically significant calls. By following the links below, you'll access the details of the phenotype calls for each center.

Phenotyping Center Significant MP Calls Pipeline
MRC HarwellBrowseIMPC_001
MRC HarwellBrowseHRWL_001
UC DavisBrowseUCD_001

Statistical Analysis

Statistical Methods

Data Statistical Method
categorical Fisher's exact test
unidimensional Wilcoxon rank sum test with continuity correction
unidimensional MM framework, generalized least squares, equation withoutWeight
unidimensional MM framework, linear mixed-effects model, equation withoutWeight

P-value distributions


Previous Releases