Explore chapters and articles related to this topic
Bootstrap and Permutation Methods
Published in Albert Vexler, Alan D. Hutson, Xiwei Chen, Statistical Testing Strategies in the Health Sciences, 2017
Albert Vexler, Alan D. Hutson, Xiwei Chen
What follows is a third more efficient iteration of the double-bootstrap resampling program now rewritten using PROC IML. What we basically did was translate the first example program into PROC IML code. The reader is referred to the SAS/IML manual for details. The data are read in exactly as before and stored in a vector called data within PROC IML. The resampled values from the inner and outer loops are stored in the vectors bootdata1 and bootdata2. The key is the resampling algorithm. It consists of first generating a random uniform number via the built-in RANUNI function and applying the function int(ranuni(0)*n)+1. This function generates a random integer from 1 to n, and facilitates sampling with replacement from the data. The rank command is specific with respect to calculating the median. The commands e_med1=sum(boot1)/brep1; e_med2=sum(boot2)/(brep1*brep2); var1=(boot1`*boot1-brep1*e_med1**2)/(brep1-1); var2=(boot2`*boot2-brep1*brep2*e_med2**2)/(brep1*brep2-1);correspond to the mean and variance calculations for , respectively, described in items 4–7 in the bootstrap algorithm overview above. data counts; label pl_ct='platelet count'; input pl_ct @@; cards; 230 222 179 191 103 293 316 520 143 226 225 255 169 204 99 107 280 226 143 259 ; proc iml; use counts; read all into data; /*Create a n by 1 vector of data counts called data*/ n=nrow(data); /*Calculate the sample size*/ bootdata1=(1:n)`; /*Initialize array for holding one bootstrap resample*/ bootdata2=(1:n)`; /*Initialize array for holding one bootstrap resample for inner loop*/ brep1=5000; /*Set the number of bootstrap resamples*/ brep2=100; /*Set the number of bootstrap resamples*/ boot1=(1:brep1)`; boot2=(1:brep1*brep2)`; in2=0; do i1=1 to brep1; do j1=1 to n; index=int(ranuni(0)*n)+1; bootdata1[j1]=data[index]; end; rankd1=rank(bootdata1); med1=0; if n/2=int(n/2) then do; /*Median for even n*/ do k1=1 to n; if rankd1[k1]=n/2 | rankd1[k1]=n/2+1 then med1=med1+bootdata1[k1]/2; end; end; else do; /*Median for odd n*/ do k1=1 to n; if rankd1[k1]=(n+1)/2 then med1=bootdata1[k1]; end; end; do i2=1 to brep2; in2=in2+1; do j2=1 to n; index=int(ranuni(0)*n)+1; bootdata2[j2]=bootdata1[index]; end; rankd2=rank(bootdata2); med2=0; if n/2=int(n/2) then do; /*Median for even n*/ do k2=1 to n; if rankd2[k2]=n/2 | rankd2[k2]=n/2+1 then med2=med2+bootdata2[k2]/2; end; end; else do; /*Median for odd n*/ do k2=1 to n; if rankd2[k2]=(n+1)/2 then med2=bootdata2[k2]; end; end; boot2[in2]=med2; end; boot1[i1]=med1; end; e_med1=sum(boot1)/brep1; e_med2=sum(boot2)/(brep1*brep2); var1=(boot1`*boot1-brep1*e_med1**2)/(brep1-1); std1=sqrt(var1); var2=(boot2`*boot2-brep1*brep2*e_med2**2)/(brep1*brep2-1); std2=sqrt(var2); print e_med1 var1 std1 e_med2 var2 std2; quit;
Adipose PD-L1 Modulates PD-1/PD-L1 Checkpoint Blockade Immunotherapy Efficacy in Breast Cancer
Published in OncoImmunology, 2018
Bogang Wu, Xiujie Sun, Harshita B. Gupta, Bin Yuan, Jingwei Li, Fei Ge, Huai-Chin Chiang, Xiaowen Zhang, Chi Zhang, Deyi Zhang, Jing Yang, Yanfen Hu, Tyler J. Curiel, Rong Li
The exact cis-acting regulatory elements that confers activation of PD-L1 expression following adipogenesis remains to be characterized. In this regard, we surveyed published ChIP-seq data for enhancer-binding proteins preadipcoytes before and after adipose differentiation in vitro. MED1 is a transcription mediator protein frequently associated with transcriptional enhancers.63 Two clusters of MED1 ChIP-seq are observed around the Cd274/Pd-l1 locus after, but not before, differentiation (Clusters 1 and 2; Supplementary Figure 6). Cluster 1 is located in the first intronic region of the Cd274/Pd-l1 locus, and Cluster 2 in an intergenic region downstream of the gene. Of note, both MED1 clusters coincide with binding peaks for PPARγ and C/EBPβ, two master transcriptional regulators of adipogenesis.64 Future work will be aimed at validating the functional role of these and other genomic regions in adipogenesis-induced Cd274/Pd-l1 transcription.
Recent progress in development of cyclin-dependent kinase 7 inhibitors for cancer therapy
Published in Expert Opinion on Investigational Drugs, 2021
Hanzhi Liang, Jintong Du, Reham M. Elhassan, Xuben Hou, Hao Fang
In addition to CDKs, CDK7 can regulate other proteins. Metastatic castration-resistant prostate cancer is a fatal disease, primarily caused by androgen receptor (AR)-driven transcriptional addiction [47]. The transcriptional coactivator MED1, which is required for AR-mediated transcription, relies on CDK7-dependent phosphorylation at Thr1457 [48]. The CDK7 covalent inhibitor THZ1 blocked the AR/MED1 co-recruitment genome-wide, passivated AR-dependent tumor growth, and reversed hyperphosphorylation of the MED1- associated enzalutamide-resistance phenotype. In vivo, CDK7 inhibition induced AR-amplified tumor regression in castration-resistant prostate cancer xenograft mouse model [48,49].
microRNAs Alterations of Myocardium and Brain Ischemia-Reperfusion Injury: Insight to Improve Infarction
Published in Immunological Investigations, 2022
Fatemeh Sabet Sarvestani, Negar Azarpira
In this line, a recent study is showed that miRNA-146a can be upregulated and exerted a protective effect against myocardial IRI by the target gene Med1 and related to the apoptosis signaling pathway via alteration of Bax/Bcl2 ratio and caspase-3 level (Zhang et al. 2019b).