Explore potential mechanism of immunotherapy resistance in LUSC

Althrough immunotherapy has revolutionized the treatment of lung squamous carcinoma, a significant proportion of patients which had high PDL1 expression showed resistance to immnotherapy. Based on gene expression profiles, we used virtual microdissection method to deconvolute the expression patterns and identify an immunosuppressive subtype which showed potential resistance to immune checkpoint blockade therapy. Then we defined this subtype as an Exhausted Immune class and developed an Exhausted Immune classifier to predict patients belonging to this class

Then we also constructed this database web app for clinical researchers to explore the mechanism of potential immunotherapy resistance at the multiomics level.


This application consists of seven functional modules including signature expression, exhausted immune classifier, somatic mutation, microRNA, methylation, clinic and chemotherapy drugs. The dedailed usage of each module was described as below.



Signature expression

In this module,user can perfomed the co-expression analysis of given gene between exhausted immune class and rest class of LUSC cohort. The KEGG pathways, GO function and target drug of this given gene was listed in table. And user can investage expression correlation of two interested genes


Exhausted immune classifier

User can upload a lung squamous carcinoma expression matrix to predict exhausted immumne class with potential resistance to immunotherapy.


Somatic mutation

In this module,we developed 3 functional panels.Landscape of mutations panel enables user to explore overall mutation landscape of LUSC cohorts selected by tumor stages or exhausted immune class and to check amino acid changes information of individual gene among the cohort.Comparative analysis panel can make user to compare two cohorts to detect differentially mutated genes. Survival analysis panel enables user to check whether the mutation of given gene is associated with prognosis.


MicroRNA

User can browse differentially expressed microRNA between exhausted immune class and rest class, and the targeted genes of microRNA. And user can also perform correlaiton analysis of microRNA and target gene and functional enrichment analysis of the targeted genes.


Methylation

All differentially methylated CpGs between exhausted immune class and rest class can be queried by user. And the correlation between CpG methylation value and expression of corresponding gene can be performed.


Clinic

User can check association of exhausted immune class with the prognosis, gender, age, and clinical stage of LUSC patients, and can also explore whether the expression of given gene is associated with these clinical variables.


Chemotherapy drugs

User can predict chemotherapy sensitivity of patients based on expression profiles.


Explore the expression alterations of the given gene between EIC and Rest class in lung squamous cell carcinoma

Please enter a gene

Late stages(stage IIA to IV)

Loading...

Early stages(stage I to II)

Loading...

KEGG pathway, GO function and target drug of the given gene.

KEGG pathway


Loading...

GO function annotation


Loading...

Drug target

Loading...


Co-expression analysis between two interested genes.

Late stages(stage IIA to IV)

Early stages(stage I to II)



Predict exhausted immune class of patients with LUSC




Note: columns correspond to samples, rows to genes.

This is a example,you can use other samples for prediction


Download
Loading...
Loading...

Explore mutation related to potential resistance by correlation with Exhausted Immune class

Summary of mutation landscape

Loading...

Oncoplot for selected mutated genes

Loading...

Somatic mutation interactions

Loading...

Gene for lollipop plot

Lollipop plot

Loading...

Cohort A:

Cohort B:

Forest Plot

Loading...

Differentially mutated genes between two cohorts

Loading...
Loading...

















Explore microRNA related to immunotherapy resistance by correlating Exhausted Immune class(EIC)



Volcano plot

Loading...

Differentially expressed microRNA between EIC and rest class

Loading...

The experimentally validated miRNA-target genes from miRTarBase

Loading...

Functional enrichment analysis

Loading...

All differentially methylated CpG sites between exhausted immune class and rest class

Loading...

Correlation between selected CpG and correponding gene

Loading...

The correlation between gene expression and clinical information




Group by expression of input gene:

Age


Gender,tumor stage distribution of high and low gene expression groups

Gender

Tumor stages

Prediction of chemocherapy sensitive using 'pRRophetic' R package

Note: columns correspond to samples, rows to genes.

This is an example: 61 LUSC patients from GSE30219

Loading...
Note: The lower the IC50, the more sensitive they are to the drug.






















































The generated data in this study can be downloaded.


Supplementary materials along with our study.

Supplementary tables



Early-stage(stage I to II) LUSC subtype information.

Early-stage Subtype file


Late-stage(stage IIA to IV) LUSC subtype information.

Late-stage Subtype file




















Contact:Minglei Yang
Email:yangmlei3@mail2.sysu.edu.cn

Li Lab @Sun Yat-sen University