# Identification and analysis of diverse programmed cell death patterns in idiopathic pulmonary fibrosis using microarray-based transcriptome profiling and single-nucleus RNA sequencing

## TL;DR

This study identifies and analyzes programmed cell death patterns in idiopathic pulmonary fibrosis using transcriptome profiling and RNA sequencing to improve diagnosis and prognosis.

## Contribution

The study introduces novel PCDI.prog and PCDI.diag signatures for predicting IPF progression and enabling early diagnosis.

## Key findings

- The PCDI.prog signature, developed using 101 machine-learning techniques, effectively predicts outcomes in IPF patients.
- Combining PCDI.prog with clinical data improves prediction of disease progression and survival rates.
- The PCDI.diag signature provides insights for early diagnosis of IPF.

## Abstract

Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive pulmonary disorder marked by the gradual substitution of lung tissue with fibrotic tissue, resulting in respiratory failure. While the precise etiology of IPF remains unclear, an increasing number of studies have indicated that programmed cell death (PCD) significantly contributes to the onset and advancement of IPF. PCD is implicated not only in the impairment of alveolar epithelial cells during fibrosis but also in the alterations of immune cells inside the fibrotic milieu. Investigating the PCD patterns offers a novel approach to the early diagnosis and prognostic evaluation of IPF.

The study utilized microarray-based transcriptome profiling and single-nucleus RNA sequencing to identify and analyze diverse PCD patterns in IPF. IPF-related genes were identified based on differential expression analysis, univariate Cox regression analysis, the “Scissor” program, and the “Findmarkers” program. A combination of machine learning was employed to develop stable predictive and diagnostic signatures associated with IPF, based on the filtered relevant genes.

The stable PCDI.prog signature was established through the integration of 101 distinct machine-learning techniques, which exhibited superior efficacy in predicting outcomes in IPF patients through the validation of multiple datasets. Integrating PCDI.prog signature with patient clinical information, such as age, gender, and GAP score, enables the prediction of disease progression rates and patient survival. Additional PCDI.diag signature can offer insights into the early diagnosis of IPF.

In summary, PCDI.prog signature and PCDI.diag signature offer critical insights for the early diagnosis, prognostic evaluation, and personalized treatment of IPF.