[00:00:00] Speaker 03:
1103, Garden Health versus Vidal.

[00:00:04] Speaker 03:
Mr. Mills?

[00:00:06] Speaker 00:
May I please the court?

[00:00:07] Speaker 00:
This appeal presents three issues.

[00:00:09] Speaker 00:
The first issue is the meaning of the predefined regions in the claims of the 743 patent.

[00:00:14] Speaker 00:
The only meaning that matters in claim construction is the meaning in the context of the patent.

[00:00:19] Speaker 00:
Without identifying any affirmative support in the intrinsic evidence, the board adopted an extrinsic construction that a predefined region is any predefined region in a genome.

[00:00:30] Speaker 00:
including an entire chromosome.

[00:00:32] Speaker 00:
But Phillips rejected exactly this approach in which the specification is consulted only after a determination is made based on an extrinsic source as to the meaning of the claim.

[00:00:44] Speaker 00:
While petitioner's expert Dr. Gabriel proposed this construction, she admitted during her deposition that she did not even look at the specification.

[00:00:54] Speaker 01:
Isn't your argument founded on the notion that chromosome comparisons inconsistent with E1 in claim one?

[00:01:09] Speaker 00:
So it is exactly correct that defining a predefined region as including an entire chromosome is incompatible with the step E1 normalization step being used.

[00:01:19] Speaker 01:
OK, but it's not inconsistent with E2, right?

[00:01:22] Speaker 00:
The E1 prong, the first avenue for detecting copy number variation in the claim, is rendered non-functional.

[00:01:31] Speaker 01:
OK, but you didn't answer my question.

[00:01:33] Speaker 01:
It's not inconsistent with E2.

[00:01:35] Speaker 00:
We have not argued that it becomes incompatible with E2.

[00:01:39] Speaker 01:
OK, but it says and or.

[00:01:40] Speaker 01:
So does this come down to the meaning of and or?

[00:01:46] Speaker 00:
The argument.

[00:01:50] Speaker 00:
When the predefined regions are not uniform or substantially uniform, then it renders the first avenue for detecting copy number variation nonfunctional.

[00:01:58] Speaker 01:
OK, but it says and or, which would mean or, include or.

[00:02:04] Speaker 00:
Right.

[00:02:04] Speaker 00:
We do not dispute that the claims include the or.

[00:02:07] Speaker 00:
We believe that the best meaning of the claims does not render nonfunctional the first avenue for detecting copy number variation in the claim.

[00:02:17] Speaker 01:
I'm not sure that I understand where that gets us if the chromosomal comparison is consistent with E2, which you seem to agree.

[00:02:27] Speaker 00:
So in any claim construction, the first task is to look at the specification, including the claims, to understand what it means in the context of this specific pad.

[00:02:37] Speaker 00:
And the specification speaks with one voice as to the meaning of the predefined regions.

[00:02:41] Speaker 00:
They tell us that predefined regions are uniform or substantially uniform.

[00:02:46] Speaker 00:
And that's the aspect that when it's abandoned, it renders the E1 prong nonfunctional, which is a large portion of the claim.

[00:02:53] Speaker 00:
In addition, the specification tells us that the predefined regions are orders of magnitude smaller than entire chromosomes.

[00:03:00] Speaker 00:
And it also tells us that they are far more numerous than the number of chromosomes.

[00:03:05] Speaker 00:
So that's what the specification tells us.

[00:03:07] Speaker 00:
Now, the question we should be asking is, what part of the specification tells us that the predefined regions are any region, including an entire chromosome?

[00:03:15] Speaker 00:
And there really is nothing.

[00:03:17] Speaker 00:
By the time the board got to the specification, they were no longer asking the question about what does the specification tell me about the invention.

[00:03:24] Speaker 00:
They were evaluating whether anything in the specification rules out the extrinsic construction that they had already adopted.

[00:03:30] Speaker 00:
They pointed to two things in the specification that they evaluated.

[00:03:34] Speaker 00:
The first thing they pointed to was the use of the words in some embodiments.

[00:03:40] Speaker 00:
And from the use of the words in some embodiments, they reason that there must be embodiments that are not expressly delineated there.

[00:03:47] Speaker 00:
We don't necessarily disagree with the board on that point.

[00:03:50] Speaker 00:
The question is, if there are implied additional embodiments that are not expressly delineated,

[00:03:55] Speaker 00:
Are those embodiments consistent with what the specification tells us about predefined regions?

[00:04:01] Speaker 00:
Or are they inconsistent with what the specification tells us about predefined regions?

[00:04:05] Speaker 00:
If you look at the specification, you won't see any place where it tells you that the predefined region is an entire chromosome, or that the predefined regions and the claims lack uniformity, or at least substantial uniformity.

[00:04:17] Speaker 00:
The second place that the board points to in the specification is where it says that the predefined regions are in a genome.

[00:04:24] Speaker 00:
Now again, we don't disagree with the board that it's necessary that the predefined regions be in a genome.

[00:04:31] Speaker 00:
But it being a necessary prerequisite to be in a genome does not mean that being in a genome is sufficient to make it a predefined region.

[00:04:39] Speaker 00:
When you step back and look at the specification, there's nothing in the specification that tells us that predefined regions in these claims are entire chromosomes.

[00:04:48] Speaker 03:
Well, there's nothing that says it is, but at column 16, line 11 to 12,

[00:04:53] Speaker 03:
when they're suggesting that you can select predefined regions in the genome, this particular embodiment that they're referring to doesn't have the no greater than 100 kb in substantially uniform limitation.

[00:05:06] Speaker 03:
And, you know, while it says there are at least 50 regions, it goes on to say there could be, there only has to be a minimum amount of predefined windows up to 10.

[00:05:16] Speaker 03:
that would include whole chromosomes.

[00:05:18] Speaker 03:
So you have an embodiment.

[00:05:20] Speaker 03:
I certainly agree with you that there are several embodiments disclosed in this patent and that some of those embodiments, because they're limited to no more than 100 and substantially uniform, wouldn't extend to an entire chromosome.

[00:05:34] Speaker 03:
But what about this other passage at column 16, which doesn't have that limitation and which seems to me to expressly allow for

[00:05:44] Speaker 03:
at column 45, line 62 to 66.

[00:05:46] Speaker 03:
It just seems to expressly allow for the possibility of a predefined region being an entire chromosome.

[00:05:54] Speaker 00:
There are, Your Honor is correct, that there are embodiments in the patent that are not the claimed predefined regions.

[00:06:00] Speaker 00:
One example is that there's a place for the specification.

[00:06:03] Speaker 03:
When you say they're not the claimed predefined regions, they're not the predefined regions the way you want us to construe the claim, but that's a circular argument.

[00:06:10] Speaker 00:
What I'm saying is there are embodiments of the claims where they're talking about windows and bins that are not the predefined regions.

[00:06:16] Speaker 00:
And there's an example where they do talk about a predefined, they use a different word, a portion or something different, where they're talking about something that you're going to pull out and you're going to ignore because it's a bad area that you don't want to be looking at.

[00:06:28] Speaker 00:
But when you read the specification and you're looking at the actual parts of it that are related to the actual claims for the issue.

[00:06:35] Speaker 03:
So would you then argue that column 45, line 62 to 66, which talks about predefined windows, i.e.

[00:06:41] Speaker 03:
regions up to 10, that that's not relevant to these claims, that that is a different and distinct embodiment that the patentee chose not to cover in these claims?

[00:06:51] Speaker 02:
Yes.

[00:06:52] Speaker 03:
Okay, but why?

[00:06:54] Speaker 03:
Like what, the words predefined regions,

[00:06:57] Speaker 03:
Here's my problem.

[00:06:58] Speaker 03:
The word predefined regions of genome includes chromosomes in its plain and ordinary meaning.

[00:07:03] Speaker 03:
So then I have to look at the specification to ascertain whether there is a disavowal of that plain and ordinary meaning.

[00:07:10] Speaker 03:
That's just how we do claim construction.

[00:07:12] Speaker 03:
And when you have some embodiments that make it clear that it couldn't include a whole chromosome, it has to be some smaller section, but other embodiments that would allow for the inclusion of a whole chromosome, how have you successfully disavowed that broader interpretation of that term?

[00:07:28] Speaker 00:
So thank you, Your Honor.

[00:07:29] Speaker 00:
And I believe your question goes to whether there's an express definition in this patent.

[00:07:34] Speaker 00:
And we're not arguing that there's an expressed definition as to the predefined regions in this patent.

[00:07:38] Speaker 03:
But there has to be, because the plain and ordinary meaning of predefined regions of a genome undisputedly includes a chromosome.

[00:07:43] Speaker 03:
So undisputedly, no one argued otherwise.

[00:07:47] Speaker 03:
So then the question is, in the context of this patent, has the patentee made clear to the world that it intended for that claim language to encompass something different in the context of these claims?

[00:07:58] Speaker 00:
So respectfully, I disagree that the plain and ordinary meaning establishes that

[00:08:02] Speaker 00:
the predefined regions are entire chromosomes?

[00:08:04] Speaker 03:
Can be.

[00:08:05] Speaker 03:
Can be.

[00:08:05] Speaker 03:
Not is.

[00:08:06] Speaker 03:
Can be.

[00:08:07] Speaker 00:
So there is no source in the specification for that construction.

[00:08:12] Speaker 00:
It's completely extrinsic.

[00:08:13] Speaker 03:
This term goes to a predefined region of a genome, correct?

[00:08:16] Speaker 00:
So it's... Yes or no?

[00:08:18] Speaker 03:
Is that what these claim terms are covering?

[00:08:20] Speaker 03:
A predefined region of a genome.

[00:08:22] Speaker 00:
There's specific predefined regions of the genome that are used in specific ways for specific purposes.

[00:08:27] Speaker 03:
If you look at the way the board construed the claims... Does the claim extend to a predefined region of a genome?

[00:08:33] Speaker 03:
I'm just wondering what the other, a predefined region of a milk carton, a predefined region of a refrigerator.

[00:08:38] Speaker 03:
Is it a predefined region of a genome?

[00:08:41] Speaker 00:
It is true that a prerequisite of the predefined region is that it's a region of the genome.

[00:08:45] Speaker 03:
That is... Is a chromosome a region within a genome?

[00:08:49] Speaker 00:
It's not a predefined region the genome is used in the claims of the 743 pattern.

[00:08:53] Speaker 03:
All right.

[00:08:56] Speaker 03:
You know, you don't help when you say stuff like that, right?

[00:08:58] Speaker 03:
You don't help your case.

[00:08:59] Speaker 03:
Because you're not explaining to me anything about why.

[00:09:01] Speaker 03:
You understand my argument perfectly.

[00:09:04] Speaker 03:
And instead, you're just kind of throwing words at me, but it doesn't actually respond to my argument and concern.

[00:09:10] Speaker 03:
And I don't know how to decide this case yet, but you're not convincing me.

[00:09:14] Speaker 00:
Yes.

[00:09:14] Speaker 00:
So I apologize if it seemed to the court that I wasn't being responsive.

[00:09:18] Speaker 00:
I'm trying to be responsive to the court's questions.

[00:09:20] Speaker 00:
The point that I'm making is that when you read the specification carefully, there are various embodiments.

[00:09:26] Speaker 00:
And some of those embodiments involve windows or bins where the regions are not predefined.

[00:09:31] Speaker 00:
When you predefine the regions, it's important that the regions be uniform or substantially uniform.

[00:09:37] Speaker 00:
Because if they're not, you don't get the ability to detect the copy number variation that you do have to predefine.

[00:09:41] Speaker 03:
The problem is that's only with regard to some embodiment.

[00:09:44] Speaker 03:
There are other embodiments in the spec that don't have that limitation.

[00:09:49] Speaker 00:
I think if your honor reads the specification carefully, the parts of the specification that discuss using the predefined regions are uniformly, substantially uniform regions.

[00:10:00] Speaker 00:
And that if they're not discussing substantial uniformity, that it's not the predefined region embodiment.

[00:10:06] Speaker 03:
What about column 45?

[00:10:06] Speaker 03:
The one that you already agreed is a different embodiment, but you argued not covered by the scope of these

[00:10:18] Speaker 00:
Yes, column 45 is not the predefined regions of the claims.

[00:10:24] Speaker 03:
Because you say so.

[00:10:26] Speaker 03:
That's not the helpful part.

[00:10:28] Speaker 00:
Well, I mean, that's not what it's talking about.

[00:10:30] Speaker 00:
It's not what it says.

[00:10:31] Speaker 00:
So for example, in column 45, it says there's predefined windows that are known throughout the genome to be hard to sequence, and therefore you filter them out from the data set.

[00:10:40] Speaker 00:
That's not the predefined regions that the claim is talking about.

[00:10:44] Speaker 00:
It's a different embodiment.

[00:10:45] Speaker 00:
They're doing something different.

[00:10:52] Speaker 00:
If I can move on to the unique sequence reads argument, the second issue on the field.

[00:11:01] Speaker 00:
With respect to claims 24 and 26, the board construed unique sequence reads to contain a unique sequence such that the sequence can be identified without the need for a barcode.

[00:11:14] Speaker 00:
In making this construction, the board pointed to column 37 of the specification, which discusses unique sequence data used for identifying individual molecules in the sample.

[00:11:25] Speaker 00:
In the institution decision, the board applied that construction and evaluated whether the true reference satisfies it.

[00:11:31] Speaker 01:
The board had looked at... In the patent owner's response, he didn't raise any different arguments with respect to claims 24 and 26, right?

[00:11:44] Speaker 00:
In the Pat Nooner response, the Pat Nooner response rested on the status quo, which were the arguments that were made pre-institution.

[00:11:51] Speaker 01:
It just didn't say anything that was different about 24 and 26, right?

[00:11:56] Speaker 00:
There wasn't an additional separate argument in the Pat Nooner response for claims 24 and 26.

[00:12:01] Speaker 01:
So why isn't there a forfeiture?

[00:12:03] Speaker 00:
It's not forfeited because the issue was live in the case.

[00:12:06] Speaker 00:
It was disputed by the parties and it was addressed by the board.

[00:12:09] Speaker 00:
The board certainly didn't think it was forfeited.

[00:12:11] Speaker 00:
They addressed it.

[00:12:12] Speaker 00:
They addressed it in a manner that changed the outcome from what it had been in the institution decision.

[00:12:18] Speaker 00:
They retained the claim construction that they had applied at institution, but they applied it differently, where before they recognized that uniqueness of the loci in the genome does not tell you whether the molecules are unique sequence rates of those molecules.

[00:12:33] Speaker 00:
So at institution they got it right.

[00:12:35] Speaker 00:
In the final written decision, they got confused about the application.

[00:12:39] Speaker 00:
And the status quo was that Garden had already explained why the board's adopted construction was not satisfied by CHU.

[00:12:49] Speaker 00:
And because the board provides no meaningful rationale under its construction for applying to CHU, we would ask that the board's judgment as to Claims 24 and 26 be reversed.

[00:13:01] Speaker 02:
Okay.

[00:13:02] Speaker 02:
Ms.

[00:13:03] Speaker 02:
Crayman?

[00:13:04] Speaker 02:
Good morning, Your Honors, and may it please the Court.

[00:13:07] Speaker 02:
The Board correctly construed predefined regions to include entire chromosomes.

[00:13:11] Speaker 02:
There's nothing in the claims that limits the predefined regions to any specific length or number, and the Board did not err in declining to read the specific embodiments into the claims.

[00:13:21] Speaker 02:
However, as Chief Judge Moore pointed out, there is an embodiment that would encompass entire chromosomes where the windows are at least 10 and at least 20, which would include

[00:13:33] Speaker 02:
the 23 chromosomes of the human chromosome.

[00:13:36] Speaker 02:
And while counsel says this is not the predefined regions that are claimed, in their motion to amend, and this is at JA 677, they particularly pointed for predefined regions for support in their earlier priority applications that they were the same as chromosomal windows.

[00:13:57] Speaker 02:
So I think it's hard for them to then say now that predefined regions and windows and bins and the other terms used in the specification aren't referring to the same embodiment.

[00:14:10] Speaker 02:
Garden has inserted that the predefined regions as entire chromosomes doesn't make sense in the context of the claim language because you can't detect a copy number variation just with E1 as in CHU.

[00:14:25] Speaker 02:
However, the board found that Chu detects a copy number variation in aneuploidy using step E1 and step E2 of Chu.

[00:14:34] Speaker 02:
And the claim, as Judge Dyke points out expressly, permits that process for detecting a copy number variation.

[00:14:42] Speaker 02:
And so there's nothing inconsistent with the claim language of claim one in using an entire chromosome as a predefined region.

[00:14:49] Speaker 03:
If there wasn't an or there and if

[00:14:51] Speaker 03:
For example, if there was a dependent claim that says this determining step has to be performed exclusively by E1 without the or option, would this be a different case?

[00:15:02] Speaker 02:
It would need to have a different record for that to be found by Chu.

[00:15:07] Speaker 02:
Chu does have a section where it does detect a copy number variation just using E1.

[00:15:13] Speaker 02:
But the board didn't rely on that in this case.

[00:15:16] Speaker 02:
It relied on the fact that there's the normalization across chromosomes.

[00:15:19] Speaker 02:
And then you go on to compare.

[00:15:22] Speaker 02:
to a reference, a control sample, to detect the copy number variation.

[00:15:26] Speaker 01:
24 and 26 require E1, right?

[00:15:31] Speaker 02:
So they require E1, but they break it down then by whether there's its reads or unique sequence reads.

[00:15:41] Speaker 02:
So it doesn't remove step the and or of 2E2 in... I guess the argument would be that with respect to 24 and

[00:15:51] Speaker 01:
26 that because you have to use E1 that that's inconsistent with a chromosomal comparison.

[00:16:03] Speaker 01:
They don't really make that argument.

[00:16:05] Speaker 02:
They don't make that argument, and I don't think they're reading claim 24 that way.

[00:16:09] Speaker 02:
I think what they're reading claim 24 to do is to limit claim 1 to just unique sequence reads as opposed to reads or unique sequence reads.

[00:16:19] Speaker 02:
So claim 24 doesn't require you to detect the copy number variation just using step E1, but it requires when you use step E1 that you're using unique sequence reads.

[00:16:30] Speaker 02:
So it's still in claim 24 and 26 has that and or.

[00:16:35] Speaker 02:
proposition for detecting the copy number variation.

[00:16:41] Speaker 02:
Before then, turning to the argument about the unique sequence reads, Garden didn't make a separate argument about claims 24 and 26 in its Patent Owned Response as the board expressly stated.

[00:16:54] Speaker 02:
The board didn't say forfeiture or waiver, but when the board says, I don't see a separate argument for patentability, that's essentially what the board means.

[00:17:03] Speaker 02:
The board then goes on to rely on the patent owner's evidence, expert testimony, that Chu does in fact teach unique sequence reads through this filtering step.

[00:17:13] Speaker 02:
And that is not inconsistent with the board's institution decision.

[00:17:17] Speaker 02:
The board's construction of unique sequence reads was really a tautology except for the use of barcodes.

[00:17:25] Speaker 02:
Unique sequence reads are reads with unique sequences

[00:17:29] Speaker 02:
that don't have to use barcodes then for you to identify them as unique sequence reads.

[00:17:34] Speaker 02:
And it's undisputed that you does not use barcodes.

[00:17:37] Speaker 02:
And in fact, the board went on for claims 24 and 26 in the set ground too with Kind, where Kind does use barcodes.

[00:17:44] Speaker 02:
And even though Garden hadn't made a separate argument for those claims, notes that under its construction of unique sequence reads, Kind can't support that obviousness decision.

[00:17:54] Speaker 02:
So it's not inconsistent with the board's construction of unique sequence reads.

[00:17:58] Speaker 02:
Gartner gives a number of new constructions in the reply brief before this court as to why Chu doesn't teach unique sequence reads, but those were not arguments that were presented to the board.

[00:18:09] Speaker 02:
And then it's also consistent with what the board found in its institution decision.

[00:18:15] Speaker 02:
It found that Chu, for claim one, uses reads, but it didn't reject the expert's testimony about the filtering to produce

[00:18:23] Speaker 02:
unique sequence reads, which is the testimony that it then relies on in its final written decision.

[00:18:29] Speaker 02:
We didn't get to the motion to amend.

[00:18:31] Speaker 02:
Council, on the other side, didn't get to the motion to amend.

[00:18:33] Speaker 02:
I'm happy to answer any questions this court may have about that.

[00:18:37] Speaker 02:
And if there are no questions, ask this court to please affirm the board's decision.

[00:18:42] Speaker 03:
Thank you.

[00:18:43] Speaker 03:
Thank you, Ms.

[00:18:44] Speaker 03:
Graydon.

[00:18:44] Speaker 03:
Mr. Mills, you have some rebuttal time.

[00:18:46] Speaker 00:
Thank you, Your Honor.

[00:18:47] Speaker 00:
I want to just address two questions that were discussed previously by my friend.

[00:18:53] Speaker 00:
So to Judge Dyck's question about claims 24 and 26 requiring E-1, I believe that they do require E-1 and that it was the burden of the petitioner to demonstrate their unpatentability, that that burden never shifted to the patent owner.

[00:19:09] Speaker 00:
And to the discussion of column 45,

[00:19:12] Speaker 00:
to Judge Moore's questions.

[00:19:14] Speaker 00:
I believe reviewing column 45, if you read it very carefully, I think the way to understand it is if you look at line 15, for instance, after the data filtering and mapping, they say the plurality of sequencers generates a chromosomal region of coverage.

[00:19:31] Speaker 00:
Then they're looking at a variety of things that you can do after you've already done this quantification and mapping step onto the chromosome.

[00:19:38] Speaker 00:
And so in some of these cases,

[00:19:40] Speaker 00:
They're setting windows or bins based on trying to get a certain number of reads in each window, right?

[00:19:45] Speaker 00:
They're doing something different.

[00:19:46] Speaker 00:
It's different than a predefined region that's set before you're doing your quantification.

[00:19:51] Speaker 00:
And I think that's the reason that this embodiment on 45 is not the claimed embodiment.

[00:19:56] Speaker 00:
Unless there are further questions, we'll ask the board for the remedies in our brief.

[00:20:01] Speaker 03:
Okay, we thank both counsels in this case to take another submission.