Skip to main content

TAU study finds widespread misinterpretation of gene expression data

Written on |

Bias can be now be removed from data to filter out false results, researchers say

Reproducibility of research data is a major challenge in experimental biology. As data generated by genomic-scale techniques increases in complexity, this concern is becoming more and more worrisome. RNA-seq, one of the most widely used methods in modern molecular biology, allows in a single test the simultaneous measurement of the expression level of all genes in a given sample. New research by a Tel Aviv University group identifies a frequent technical bias in data generated by RNA-seq technology, which often leads to false results. The study was conducted by Dr. Shir Mandelbaum, Dr. Zohar Manber, Dr. Orna Elroy-Stein and Dr. Ran Elkon at TAU’s Sackler Faculty of Medicine and George S. Wise Faculty of Life Sciences and was published on November 12 in PLOS Biology. “Recent years have witnessed a growing alarm about false results in biological research, sometimes referred to as the reproducibility crisis,” Dr. Elkon, lead author of the study, says. “This study emphasizes the importance of proper statistical handling of data to lessen the number of misleading findings.”

Removing bias from science

A main goal of RNA-seq experiments is to characterize biological processes that are activated or repressed in response to different conditions. The researchers analyzed dozens of publicly available RNA-seq datasets to profile the cellular responses to numerous stresses. During the research, the scientists noticed that sets of particularly short or long genes repeatedly showed changes in the expression level measured by the apparent number of RNA transcripts from a given gene. Puzzled by this recurring pattern, the team wondered whether it reflected some universal biological response common to different triggers or whether it stemmed from some experimental condition. To tackle this question, they compared replicated samples from the same biological condition. Differences in gene expression between replicates can reflect technical effects that are not related to the experiment’s biological factor of interest. Unexpectedly, the same pattern of particularly short or long genes showing changes in expression level was observed in these comparisons between replicates. This pattern is the result of a technical bias that seemed to be coupled with gene length, the researchers say. Importantly, the TAU researchers were able to show how the length bias they detected in many RNA-seq datasets led to the false identification of specific biological functions as cellular responses to the conditions tested. “Such misinterpretation of the data could lead to completely misleading conclusions,” Dr. Elkon concludes. “In practical terms, the study also shows how this bias can be removed from the data, thus filtering out false results while preserving the biologically relevant ones.”

Related posts

Breaking the MedTech Glass Ceiling

27 March 2024

Summer Glow: How Sun Exposure Boosts Fertility in Women Ages 30-40

7 March 2024

Are We Close to Ending Alzheimer’s Memory Loss?

15 February 2024

Destroying Cancer: new drug delivery system containing RNA therapy can target cancer cells in bone marrow

31 July 2023

Stress Makes Vaccines Less Effective

26 July 2023

Researchers Induce Cancer Cell “Suicide”

17 July 2023

Metabolomics – A New Frontier in Preventive Medicine

13 July 2023

Older Bats do Suffer from Age-related Hearing Loss

13 July 2023

Operation Guardian of the Walls: Women, Young People and Residents of the South Paid the Heaviest Price

12 July 2023

Discovery May Lead to Personalized Medicine for Infectious Diseases

12 July 2023

One Third of Normal-Weight Individuals are Obese

12 July 2023

Breakthrough Gene Therapy Offers Hope for Severe Developmental Epilepsy in Children

27 June 2023

Prof. Isaac P. Witz Honored with 2023 Szent-Györgyi Prize for Progress in Cancer Research

26 June 2023

A Step Closer to Beating Melanoma?

18 June 2023

Imagining the Future Patient

13 June 2023

New Treatment Reduces ADHD Symptoms in 1 out of 3 Students

13 June 2023

Victoria

Tok Corporate Centre, Level 1,
459 Toorak Road, Toorak VIC 3142
Phone: +61 3 9296 2065
Email: [email protected]

New South Wales

Level 22, Westfield Tower 2, 101 Grafton Street, Bondi Junction NSW 2022
Phone: +61 418 465 556
Email: [email protected]

Western Australia

P O Box 36, Claremont,
WA  6010
Phone: :+61 411 223 550
Email: [email protected]