bedClip is a command line program to do this:

bedClip input.bed http://hgdownload.cse.ucsc.edu/goldenPath/hg38/bigZips/hg38.chrom.sizes output.bed

You can download bedClip from the directory appropriate to your operating system within our directory of utilities.

If you have questions about running bedClip, feel free to send a question to one of our mailing lists:

• genome@soe.ucsc.edu for general questions
• genome-mirror@soe.ucsc.edu for questions involving mirrors or gbibs
• genome-www@soe.ucsc.edu for questions involving private data

ChrisL from the UCSC Genome Browser

generate a awk script that will clip your bed records:

mysql --user=genome -N --host=genome-mysql.cse.ucsc.edu -A -D hg19  -e 'select chrom,size from chromInfo '  |\
awk '{printf("($1==\"%s\") {L=%d;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf(\"%s\\t%%d\\t%%d\\n\",B,E);next;}\n",$1,$2,$1);}' > script.awk



$ head  script.awk
($1=="chr1") {L=249250621;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf("chr1\t%d\t%d\n",B,E);next;}
($1=="chr2") {L=243199373;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf("chr2\t%d\t%d\n",B,E);next;}
($1=="chr3") {L=198022430;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf("chr3\t%d\t%d\n",B,E);next;}
($1=="chr4") {L=191154276;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf("chr4\t%d\t%d\n",B,E);next;}
($1=="chr5") {L=180915260;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf("chr5\t%d\t%d\n",B,E);next;}
($1=="chr6") {L=171115067;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf("chr6\t%d\t%d\n",B,E);next;}
($1=="chr7") {L=159138663;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf("chr7\t%d\t%d\n",B,E);next;}
($1=="chrX") {L=155270560;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf("chrX\t%d\t%d\n",B,E);next;}
($1=="chr8") {L=146364022;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf("chr8\t%d\t%d\n",B,E);next;}
($1=="chr9") {L=141213431;B=int($2);E=int($3);B=(B>=L?L:B);E=(E>=L?L:E);printf("chr9\t%d\t%d\n",B,E);next;}

then use this awk script :

awk -f  script.awk input.bed

To solve this problem more generically, make a BED file and use that as a mask with BEDOPS bedops --element-of:

$ fetchChromSizes hg38 | awk '{ print $1"\t0\t"$2; }" | sort-bed - > hg38.bed
$ bedops --element-of 1 in.bedGraph hg38.bed > masked.in.bedGraph

Then convert the masked bedGraph file to Wiggle format.

But mainly I'd be concerned about having signal get generated in regions that don't or shouldn't exist. That might point to a potential data problem or code smell, somewhere.