CheXpert is a dataset consisting of 224,316 chest radiographs of 65,240 patients who underwent a radiographic examination at Stanford Health Care between October 2002 and July 2017, in both inpatient and outpatient centers. Included are their associated radiology reports.

This file contains a mini-website of Tables S1-S10. The same mini-website is available at http://carterdev.jax.org/dtx/a2/index.html. Interactive expression plots. Table S1. Number and proportion of isolated cells for cytological analysis and RNA-seq. Table S2. Gene lists and gene expression values for each of the substages, separated by worksheets. An additional worksheet provides results for X-linked genes. Abbreviations for each substage are as follows: Sp’gonia for spermatogonia, PL for preleptotene, EL for early leptotene, LL + Z for late leptotene and zygotene, LL + Z + EP for those genes found in late leptotene, zygotene, and early pachytene, EP for early pachytene, EP + LP + D for early pachytene, late pachytene, and diplotene, and LP + D for late pachytene and diplotene. Sheet names prefixed with “Anti-“correspond to genes that are negatively concordant with the substage, (e.g., Anti-PL corresponds to genes that are negatively concordant with preleptotene. Genes are ranked by how closely the expression pattern follows the substage pattern (rank = 1 is best). Gene ids and Gene names for piRNA precursors and NONCODE lncRNAs have prefix “pi-” and “NON”, respectively. Expression is given in log2(TPM + 1). Table S3. GO analysis for substage-negatively concordant genes. Abbreviations for each substage are as follows: Sp’gonia for spermatogonia, EL for early leptotene, LL-Z for late leptotene and zygotene, EP for early pachytene, and LP-D for late pachytene and diplotene. Table S4. GO analysis for substage-concordant genes. Abbreviations for each substage are as follows: Sp’gonia for spermatogonia, EL for early leptotene, LL-Z for late leptotene and zygotene, EP for early pachytene, and LP-D for late pachytene and diplotene. Table S5. Analysis of genes annotated to meiosis. Table S6. TFs for substage-concordant genes. The column names are explained as follows: “NES” is short for Normalized Enrichment Score in iRegulon, “ngenes” is the number of substage-concordant genes associated with the TF, “name” used in the arc-diagram (* indicates a family of genes, e.g., Rfx* for Rfx1, Rfx2, Rfx3, Rfx4). Bolded TFs are highly expressed in the testis and those highlighted in red are associated with male reproduction. Table S7. TFs for substage-negatively concordant genes. The column names are explained as follows: “NES” is short for Normalized Enrichment Score in iRegulon, “ngenes” is the number of substage negatively-concordant genes associated with the TF, “name” used in the arc-diagram (* indicates a family of genes, e.g., Rfx* for Rfx1, Rfx2, Rfx3, Rfx4). Table S8. Comparison of substage-concordant gene lists with derived clusters in Margolin, et al. [14]. Table S9. Comparison of substage-concordant gene lists with sets derived from FACS-sorted cells [5, 6]. Table S10. Primers for qRT-PCR. (ZIP 6513 kb)

This file contains a mini-website of Figures S1-S6 and a link to interactive expression plots. The same mini-website is available at http://carterdev.jax.org/dtx/a1/index.html. Figure S1. Expression of substage-concordant genes. (A-H) Heat maps show gene expression of substage-concordant genes at each time point. Gene expression is shown as average log2(TPM + 1) of time point replicates. Figure S2. Expression of substage-negatively concordant genes. (A-H) Heat maps show gene expression of substage-negatively concordant genes at each time point. Gene expression is shown as average log2(TPM + 1) of time point replicates. Figure S3. Validation of RNA-seq expression pattern by qRT-PCR. Box plots (top) show gene expression detected by RNA-seq at 10 dpp and 16 dpp. Bar graphs (bottom) show gene expression detected by qRT-PCR using isolated germ cells from pooled testes from 3 mice at 10 dpp and 16 dpp. Gene expression for RNA-seq is shown as log2(TPM + 1). Gene expression for qRT-PCR is shown as fold increase relative to Actb expression. Figure S4. TF arc diagram illustrating the most significant TFs across substages for the concordant gene lists. Node color represents the substage. Node size is related to significance (larger = more significant but all are highly significant with a iRegulon NES > = 4). Width of arc corresponds to how many TFs are shared between nodes (wider = more shared TFs). A node is connected to another node if a node’s TFs are a subset of the other node’s TFs. Color of the arc (degree) is related to how many connections the node has. Degree equals the number of connections. The name of each node is an abbreviation, where an asterisk indicates multiple members of the TF family are included in the node (e.g., Rfx* for Rfx1, Rfx2, Rfx3, Rfx4) and when a node name is followed by + integer, as in “Mef2* + 1”, it indicates that there are TFs in the Mef2 family plus one other TF. All TFs associated with each node are listed in Additional file 1: Table S6. The number in parenthesis next to the node name indicates the number of substage-concordant genes associated with the TF cluster. Figure S5. TF arc diagram illustrating the most significant TFs across substages for the negatively concordant gene lists. Node color represents the substage. Node size is related to significance (larger = more significant but all are highly significant with a iRegulon NES > = 4). Width of arc corresponds to how many TFs are shared between nodes (wider = more shared TFs). A node is connected to another node if a node’s TFs are a subset of the other node’s TFs. Color of the arc (degree) is related to how many connections the node has. Degree equals the number of connections. The name of each node is an abbreviation, where an asterisk indicates multiple members of the TF family are included in the node (e.g., Rfx* for Rfx1, Rfx2, Rfx3, Rfx4) and when a node name is followed by + integer, as in “Mef2* + 1”, it indicates that there are TFs in the Mef2 family plus one other TF. All TFs associated with each node are listed in Additional file 1: Table S7. The number in parenthesis next to the node name indicatea the number of substage-negatively-concordant genes associated with the TF cluster. Figure S6. Expression of TF and its target genes for each substage. The difference of mean gene expression at 8 dpp of the TF and its target genes at each time point. To illustrate the overall pattern, a smoothed line was fit to the substage-specific gene expression. Color of lines represents the substage of target genes. Gene expression is shown as average log2(TPM + 1) across time point replicates. (ZIP 13261 kb)

This file contains a mini-website of Figures S1-S6 and a link to interactive expression plots. The same mini-website is available at http://carterdev.jax.org/dtx/a1/index.html. Figure S1. Expression of substage-concordant genes. (A-H) Heat maps show gene expression of substage-concordant genes at each time point. Gene expression is shown as average log2(TPM + 1) of time point replicates. Figure S2. Expression of substage-negatively concordant genes. (A-H) Heat maps show gene expression of substage-negatively concordant genes at each time point. Gene expression is shown as average log2(TPM + 1) of time point replicates. Figure S3. Validation of RNA-seq expression pattern by qRT-PCR. Box plots (top) show gene expression detected by RNA-seq at 10 dpp and 16 dpp. Bar graphs (bottom) show gene expression detected by qRT-PCR using isolated germ cells from pooled testes from 3 mice at 10 dpp and 16 dpp. Gene expression for RNA-seq is shown as log2(TPM + 1). Gene expression for qRT-PCR is shown as fold increase relative to Actb expression. Figure S4. TF arc diagram illustrating the most significant TFs across substages for the concordant gene lists. Node color represents the substage. Node size is related to significance (larger = more significant but all are highly significant with a iRegulon NES > = 4). Width of arc corresponds to how many TFs are shared between nodes (wider = more shared TFs). A node is connected to another node if a node’s TFs are a subset of the other node’s TFs. Color of the arc (degree) is related to how many connections the node has. Degree equals the number of connections. The name of each node is an abbreviation, where an asterisk indicates multiple members of the TF family are included in the node (e.g., Rfx* for Rfx1, Rfx2, Rfx3, Rfx4) and when a node name is followed by + integer, as in “Mef2* + 1”, it indicates that there are TFs in the Mef2 family plus one other TF. All TFs associated with each node are listed in Additional file 1: Table S6. The number in parenthesis next to the node name indicates the number of substage-concordant genes associated with the TF cluster. Figure S5. TF arc diagram illustrating the most significant TFs across substages for the negatively concordant gene lists. Node color represents the substage. Node size is related to significance (larger = more significant but all are highly significant with a iRegulon NES > = 4). Width of arc corresponds to how many TFs are shared between nodes (wider = more shared TFs). A node is connected to another node if a node’s TFs are a subset of the other node’s TFs. Color of the arc (degree) is related to how many connections the node has. Degree equals the number of connections. The name of each node is an abbreviation, where an asterisk indicates multiple members of the TF family are included in the node (e.g., Rfx* for Rfx1, Rfx2, Rfx3, Rfx4) and when a node name is followed by + integer, as in “Mef2* + 1”, it indicates that there are TFs in the Mef2 family plus one other TF. All TFs associated with each node are listed in Additional file 1: Table S7. The number in parenthesis next to the node name indicatea the number of substage-negatively-concordant genes associated with the TF cluster. Figure S6. Expression of TF and its target genes for each substage. The difference of mean gene expression at 8 dpp of the TF and its target genes at each time point. To illustrate the overall pattern, a smoothed line was fit to the substage-specific gene expression. Color of lines represents the substage of target genes. Gene expression is shown as average log2(TPM + 1) across time point replicates. (ZIP 13261 kb)

This file contains a mini-website of Tables S1-S10. The same mini-website is available at http://carterdev.jax.org/dtx/a2/index.html. Interactive expression plots. Table S1. Number and proportion of isolated cells for cytological analysis and RNA-seq. Table S2. Gene lists and gene expression values for each of the substages, separated by worksheets. An additional worksheet provides results for X-linked genes. Abbreviations for each substage are as follows: Sp’gonia for spermatogonia, PL for preleptotene, EL for early leptotene, LL + Z for late leptotene and zygotene, LL + Z + EP for those genes found in late leptotene, zygotene, and early pachytene, EP for early pachytene, EP + LP + D for early pachytene, late pachytene, and diplotene, and LP + D for late pachytene and diplotene. Sheet names prefixed with “Anti-“correspond to genes that are negatively concordant with the substage, (e.g., Anti-PL corresponds to genes that are negatively concordant with preleptotene. Genes are ranked by how closely the expression pattern follows the substage pattern (rank = 1 is best). Gene ids and Gene names for piRNA precursors and NONCODE lncRNAs have prefix “pi-” and “NON”, respectively. Expression is given in log2(TPM + 1). Table S3. GO analysis for substage-negatively concordant genes. Abbreviations for each substage are as follows: Sp’gonia for spermatogonia, EL for early leptotene, LL-Z for late leptotene and zygotene, EP for early pachytene, and LP-D for late pachytene and diplotene. Table S4. GO analysis for substage-concordant genes. Abbreviations for each substage are as follows: Sp’gonia for spermatogonia, EL for early leptotene, LL-Z for late leptotene and zygotene, EP for early pachytene, and LP-D for late pachytene and diplotene. Table S5. Analysis of genes annotated to meiosis. Table S6. TFs for substage-concordant genes. The column names are explained as follows: “NES” is short for Normalized Enrichment Score in iRegulon, “ngenes” is the number of substage-concordant genes associated with the TF, “name” used in the arc-diagram (* indicates a family of genes, e.g., Rfx* for Rfx1, Rfx2, Rfx3, Rfx4). Bolded TFs are highly expressed in the testis and those highlighted in red are associated with male reproduction. Table S7. TFs for substage-negatively concordant genes. The column names are explained as follows: “NES” is short for Normalized Enrichment Score in iRegulon, “ngenes” is the number of substage negatively-concordant genes associated with the TF, “name” used in the arc-diagram (* indicates a family of genes, e.g., Rfx* for Rfx1, Rfx2, Rfx3, Rfx4). Table S8. Comparison of substage-concordant gene lists with derived clusters in Margolin, et al. [14]. Table S9. Comparison of substage-concordant gene lists with sets derived from FACS-sorted cells [5, 6]. Table S10. Primers for qRT-PCR. (ZIP 6513 kb)