Final JOSS: fixed track figure, section reorder, all 4 authors, enlarged logo

Author: SarojaSomu

README.md

@@ -1,7 +1,7 @@
 # PyPeakRankR
 
 <p align="center">
-  <img src="PyPR.png" width="180" alt="PyPeakRankR logo"/>
+  <img src="PyPR.png" width="320" alt="PyPeakRankR logo"/>
 </p>
 
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
@@ -223,6 +223,10 @@ Saroja Somasundaram — Allen Institute for Brain Science
 
 Nelson J. Johansen — Allen Institute for Brain Science
 
+Jeremy Miller — Allen Institute for Brain Science
+
+Trygve E. Bakken — Allen Institute for Brain Science
+
 ## License
 
 MIT License. See [LICENSE](LICENSE).

paper/biccn_three_peaks.png

@@ -62,17 +62,6 @@ The target audience is computational biologists who work with single-cell ATAC-s
 bulk ATAC-seq data and need to systematically prioritize peaks for experimental
 follow-up, particularly for enhancer discovery or adeno-associated virus (AAV) tool design.
 
-Existing tools address related but distinct problems: peak callers such as
-MACS2 [@Zhang2008] identify open chromatin regions but rank peaks only by
-fold change or p-value, which reflects signal strength rather than cell-type
-specificity. A peak with high MACS2 fold change may be active across many cell
-types (a housekeeping element) and therefore a poor candidate for cell-type
-targeted AAV tools. Differential accessibility tools such as ArchR [@Corces2018]
-test for cell type enrichment but operate within their own data model. Annotation
-tools such as GREAT [@McLean2010] link peaks to genes. None provide a unified,
-flexible framework for assembling a standardized feature matrix across
-heterogeneous input tracks — which is precisely what PyPeakRankR addresses.
-
 # State of the field
 
 Several tools perform individual aspects of peak level feature computation.
@@ -84,6 +73,17 @@ but does not produce portable, tool agnostic feature tables. `pyfaidx`
 [@Shirley2015] enables FASTA sequence access but provides no genomics feature
 pipeline.
 
+Existing tools address related but distinct problems: peak callers such as
+MACS2 [@Zhang2008] identify open chromatin regions but rank peaks only by
+fold change or p-value, which reflects signal strength rather than cell-type
+specificity. A peak with high MACS2 fold change may be active across many cell
+types (a housekeeping element) and therefore a poor candidate for cell-type
+targeted AAV tools. Differential accessibility tools such as ArchR [@Corces2018]
+test for cell type enrichment but operate within their own data model. Annotation
+tools such as GREAT [@McLean2010] link peaks to genes. None provide a unified,
+flexible framework for assembling a standardized feature matrix across
+heterogeneous input tracks — which is precisely what PyPeakRankR addresses.
+
 PyPeakRankR fills this gap by combining `pyBigWig`, `pyfaidx`, and
 `scipy` [@Virtanen2020] into a CLI pipeline that assembles heterogeneous
 features into a single reproducible TSV table.

paper/paper.md

@@ -11,7 +11,9 @@ requires-python = ">=3.9"
 license = {file = "LICENSE"}
 authors = [
   {name = "Saroja Somasundaram", email = "sarojas@alleninstitute.org"},
-  {name = "Nelson J. Johansen", email = "nelsonj@alleninstitute.org"}
+  {name = "Nelson J. Johansen", email = "nelsonj@alleninstitute.org"},
+  {name = "Jeremy Miller", email = "jeremym@alleninstitute.org"},
+  {name = "Trygve E. Bakken", email = "trygveb@alleninstitute.org"}
 ]
 keywords = [
   "genomics", "ATAC-seq", "BigWig", "regulatory elements",